INTEL 290688-001

R
Intel 815 Chipset Family: 82815
Graphics and Memory Controller
Hub (GMCH)
Datasheet
June 2000
Document Reference Number: 290688-001
82815 GMCH
R
Information in this document is provided in connection with Intel products. No license, express or implied, by estoppel or otherwise, to any intellectual
property rights is granted by this document. Except as provided in Intel’s Terms and Conditions of Sale for such products, Intel assumes no liability
whatsoever, and Intel disclaims any express or implied warranty, relating to sale and/or use of Intel products including liability or warranties relating to fitness
for a particular purpose, merchantability, or infringement of any patent, copyright or other intellectual property right. Intel products are not intended for use in
medical, life saving, or life sustaining applications.
Intel may make changes to specifications and product descriptions at any time, without notice.
Designers must not rely on the absence or characteristics of any features or instructions marked "reserved" or "undefined." Intel reserves these for future
definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them.
The Intel 815 chipset GMCH may contain design defects or errors known as errata which may cause the product to deviate from published
specifications. Current characterized errata are available on request.
Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order.
2
2
I C is a 2-wire communications bus/protocol developed by Philips. SMBus is a subset of the I C bus/protocol and was developed by Intel. Implementations
of the I2C bus/protocol may require licenses from various entities, including Philips Electronics N.V. and North American Philips Corporation.
Alert on LAN is a result of the Intel-IBM Advanced Manageability Alliance and a trademark of IBM
Copies of documents which have an ordering number and are referenced in this document, or other Intel literature, may be obtained from:
Intel Corporation
www.intel.com
or call 1-800-548-4725
*Third-party brands and names are the property of their respective owners.
Copyright © Intel Corporation 2000
2
Datasheet
82815 GMCH
R
Contents
1.
Overview.....................................................................................................................................13
1.1.
1.2.
1.3.
1.4.
1.5.
1.6.
1.7.
1.8.
1.9.
1.10.
2.
Signal Description.......................................................................................................................21
2.1.
2.2.
2.3.
2.4.
2.5.
2.6.
2.7.
2.8.
2.9.
2.10.
2.11.
3.
Host Interface Signals ....................................................................................................22
System Memory Interface Signals .................................................................................23
AGP Interface Signals....................................................................................................24
2.3.1.
AGP Addressing Signals..............................................................................24
2.3.2.
AGP Flow Control Signals............................................................................25
2.3.3.
AGP Status Signals .....................................................................................25
2.3.4.
AGP Clocking Signals (Strobes) ..................................................................26
2.3.5.
AGP FRAME# Signals .................................................................................27
Display Cache Interface Signals ....................................................................................29
Hub Interface Signals.....................................................................................................30
Display Interface Signals................................................................................................30
Digital Video Output Signals/TV-Out Pins......................................................................31
Power Signals ................................................................................................................32
Clock Signals .................................................................................................................32
GMCH Power-Up/Reset Strap Options..........................................................................33
Multiplexed Display Cache and AGP Signal Mapping....................................................34
2.11.1.
Display Cache Mapping at the AGP Connector ...........................................35
Configuration Registers ..............................................................................................................37
3.1.
3.2.
3.3.
Datasheet
Related Documents .......................................................................................................13
The Intel 815 Chipset Family........................................................................................14
82815 GMCH Overview .................................................................................................16
Host Interface.................................................................................................................17
System Memory Interface ..............................................................................................17
Multiplexed AGP and Display Cache Interface ..............................................................18
1.6.1.
AGP Interface ..............................................................................................18
1.6.2.
Display Cache Interface...............................................................................18
Hub Interface..................................................................................................................18
82815 GMCH Integrated Graphics Support...................................................................19
1.8.1.
Display, Digital Video Out, and LCD/Flat Panel/Digital CRT........................19
System Clocking ............................................................................................................20
GMCH Power Delivery ...................................................................................................20
Register Nomenclature and Access Attributes ..............................................................37
PCI Configuration Space Access ...................................................................................38
3.2.1.
PCI Bus Configuration Mechanism ..............................................................38
3.2.2.
Logical PCI Bus #0 Configuration Mechanism.............................................39
3.2.3.
Primary PCI (PCI0) and Downstream Configuration Mechanism ................39
3.2.4.
Internal Graphics Device Configuration Mechanism....................................39
3.2.5.
GMCH Register Introduction........................................................................39
I/O Mapped Registers ....................................................................................................40
3.3.1.
CONF_ADDRConfiguration Address Register .........................................40
3.3.2.
CONF_DATAConfiguration Data Register ...............................................41
3
82815 GMCH
R
3.4.
3.5.
4
Host-Hub Interface Bridge/DRAM Controller Device Registers (Device 0) ................... 42
3.4.1.
VID—Vendor Identification Register (Device 0) .......................................... 44
3.4.2.
DID—Device Identification Register (Device 0)........................................... 44
3.4.3.
PCICMD—PCI Command Register (Device 0) ........................................... 45
3.4.4.
PCISTS—PCI Status Register (Device 0) ................................................... 46
3.4.5.
RID—Revision Identification Register (Device 0) ........................................ 47
3.4.6.
SUBC—Sub-Class Code Register (Device 0) ............................................. 47
3.4.7.
BCC—Base Class Code Register (Device 0).............................................. 47
3.4.8.
MLT—Master Latency Timer Register (Device 0) ....................................... 48
3.4.9.
HDR—Header Type Register (Device 0)..................................................... 48
3.4.10.
APBASE—Aperture Base Configuration Register (Device 0: AGP
Mode Only) .................................................................................................. 48
3.4.11.
SVID—Subsystem Vendor Identification Register (Device 0) ..................... 50
3.4.12.
SID—Subsystem Identification Register (Device 0) .................................... 50
3.4.13.
CAPPTR—Capabilities Pointer (Device 0) .................................................. 50
3.4.14.
GMCHCFG—GMCH Configuration Register (Device 0) ............................. 51
3.4.15.
APCONT—Aperture Control (Device 0) ...................................................... 53
3.4.16.
DRP—DRAM Row Population Register (Device 0)..................................... 54
3.4.17.
DRAMT—DRAM Timing Register (Device 0).............................................. 55
3.4.18.
DRP2—DRAM Row Population Register 2 (Device 0)................................ 56
3.4.19.
FDHC—Fixed DRAM Hole Control Register (Device 0).............................. 57
3.4.20.
PAM—Programmable Attributes Map Registers (Device 0)........................ 57
3.4.21.
SMRAM—System Management RAM Control Register (Device 0) ........... 62
3.4.22.
MISCC—Miscellaneous Control Register (Device 0) .................................. 64
3.4.23.
CAPID—Capability Identification (Device 0: AGP Mode Only).................... 66
3.4.24.
BUFF_SC—System Memory Buffer Strength Control Register
(Device 0) .................................................................................................... 67
3.4.25.
BUFF_SC2—System Memory Buffer Strength Control Register 2
(Device 0) .................................................................................................... 70
3.4.26.
SM_RCOMP—System Memory R Compensation Control Register
(Device 0) .................................................................................................... 71
3.4.27.
SM—System Memory Control Register....................................................... 72
3.4.28.
ACAPID—AGP Capability Identifier Register (Device 0: AGP
Mode Only) .................................................................................................. 73
3.4.29.
AGPSTAT—AGP Status Register (Device 0: AGP Mode Only) ................. 74
3.4.30.
AGPCMD—AGP Command Register (Device 0: AGP Mode Only) ............ 75
3.4.31.
AGPCTRL—AGP Control Register (Device 0: AGP Mode Only) ................ 76
3.4.32.
APSIZE—Aperture Size (Device 0: AGP Mode Only) ................................. 77
3.4.33.
ATTBASE—Aperture Translation Table Base Register (Device 0: AGP
Mode Only) .................................................................................................................... 78
3.4.34.
AMTT—AGP Multi-Transaction Timer (Device 0: AGP Mode Only) .......... 79
3.4.35.
LPTT—AGP Low Priority Transaction Timer Register (Device 0: AGP
Mode Only) .................................................................................................. 80
3.4.36.
GMCHCFG—GMCH Configuration Register (Device 0: AGP
Mode Only) .................................................................................................. 81
3.4.37.
ERRCMD—Error Command Register (Device 0: AGP Mode Only)........... 82
AGP/PCI Bridge Registers (Device 1: Visible in AGP Mode Only)............................... 84
3.5.1.
VID1—Vendor Identification Register (Device 1) ........................................ 85
3.5.2.
DID1—Device Identification Register (Device 1)......................................... 85
3.5.3.
PCICMD1—PCI-PCI Command Register (Device 1) .................................. 85
3.5.4.
PCISTS1—PCI-PCI Status Register (Device 1).......................................... 87
3.5.5.
RID1—Revision Identification Register (Device 1) ...................................... 88
3.5.6.
SUBC1—Sub-Class Code Register (Device 1) ........................................... 88
3.5.7.
BCC1—Base Class Code Register (Device 1)............................................ 88
Datasheet
82815 GMCH
R
3.6.
3.7.
3.8.
Datasheet
3.5.8.
MLT1—Master Latency Timer Register (Device 1) .....................................89
3.5.9.
HDR1—Header Type Register (Device 1) ...................................................89
3.5.10.
PBUSN—Primary Bus Number Register (Device 1)....................................89
3.5.11.
SBUSN—Secondary Bus Number Register (Device 1) ...............................90
3.5.12.
SUBUSN—Subordinate Bus Number Register (Device 1) ..........................90
3.5.13.
SMLT—Secondary Master Latency Timer Register (Device 1) ...................91
3.5.14.
IOBASE—I/O Base Address Register (Device 1) ........................................92
3.5.15.
IOLIMIT—I/O Limit Address Register (Device 1).........................................93
3.5.16.
SSTS—Secondary PCI-PCI Status Register (Device 1)..............................94
3.5.17.
MBASE—Memory Base Address Register (Device 1).................................95
3.5.18.
MLIMIT—Memory Limit Address Register (Device 1) .................................96
3.5.19.
PMBASE—Prefetchable Memory Base Address Register (Device 1) .........97
3.5.20.
PMLIMIT—Prefetchable Memory Limit Address Register (Device 1)..........98
3.5.21.
BCTRL—PCI-PCI Bridge Control Register (Device 1).................................99
3.5.22.
ERRCMD1—Error Command Register (Device 1) ....................................101
Graphics Device Registers (Device 2: VISIBLE IN GFX Mode Only) .........................102
3.6.1.
VID2—Vendor Identification Register (Device 2).......................................103
3.6.2.
DID2—Device Identification Register (Device 2) .......................................103
3.6.3.
PCICMD2—PCI Command Register (Device 2)........................................104
3.6.4.
PCISTS2—PCI Status Register (Device 2) ...............................................105
3.6.5.
RID2—Revision Identification Register (Device 2) ....................................106
3.6.6.
PI—Programming Interface Register (Device 2) .......................................106
3.6.7.
SUBC2—Sub-Class Code Register (Device 2) .........................................106
3.6.8.
BCC2—Base Class Code Register (Device 2) ..........................................107
3.6.9.
CLS—Cache Line Size Register (Device 2) ..............................................107
3.6.10.
MLT2—Master Latency Timer Register (Device 2) ...................................107
3.6.11.
HDR2—Header Type Register (Device 2) .................................................108
3.6.12.
BIST—BIST Register (Device 2)................................................................108
3.6.13.
GMADR—Graphics Memory Range Address Register (Device 2)...........109
3.6.14.
MMADR—Memory Mapped Range Address Register (Device 2) .............110
3.6.15.
SVID—Subsystem Vendor Identification Register (Device 2)....................110
3.6.16.
SID—Subsystem Identification Register (Device 2)...................................111
3.6.17.
ROMADR—Video BIOS ROM Base Address Register (Device 2) ............111
3.6.18.
CAPPOINT—Capabilities Pointer Register (Device 2) ..............................111
3.6.19.
INTRLINE—Interrupt Line Register (Device 2) ..........................................112
3.6.20.
INTRPIN—Interrupt Pin Register (Device 2)..............................................112
3.6.21.
MINGNT—Minimum Grant Register (Device 2).........................................112
3.6.22.
MAXLAT—Maximum Latency Register (Device 2)....................................112
3.6.23.
PM_CAPID—Power Management Capabilities ID Register (Device 2).....113
3.6.24.
PM_CAP—Power Management Capabilities Register (Device 2) .............113
3.6.25.
PM_CS—Power Management Control/Status Register (Device 2) ..........114
Display Cache Interface ...............................................................................................115
3.7.1.
DRT—DRAM Row Type ............................................................................115
3.7.2.
DRAMCL—DRAM Control Low .................................................................116
3.7.3.
DRAMCH—DRAM Control High ................................................................117
Display Cache Detect and Diagnostic Registers..........................................................118
3.8.1.
GRX—GRX Graphics Controller Index Register .......................................118
3.8.2.
MSRMiscellaneous Output.....................................................................119
3.8.3.
GR06Miscellaneous Register.................................................................119
3.8.4.
GR10Address Mapping ..........................................................................120
3.8.5.
GR11Page Selector ...............................................................................120
5
82815 GMCH
R
4.
Functional Description.............................................................................................................. 121
4.1.
4.2.
4.3.
4.4.
4.5.
4.6.
4.7.
4.8.
4.9.
4.10.
4.11.
6
System Address Map .................................................................................................. 121
4.1.1.
Memory Address Ranges .......................................................................... 122
4.1.2.
Compatibility Area ..................................................................................... 123
4.1.3.
Extended Memory Area............................................................................. 125
4.1.3.1.
System Management Mode (SMM) Memory Range .................... 128
Memory Shadowing ..................................................................................................... 129
I/O Address Space ...................................................................................................... 129
4.3.1.
GMCH Decode Rules and Cross-Bridge Address Mapping...................... 129
4.3.2.
Address Decode Rules.............................................................................. 130
4.3.2.1.
AGP Interface Decode Rules........................................................ 131
4.3.2.2.
Legacy VGA Ranges .................................................................... 132
Host Interface .............................................................................................................. 133
4.4.1.
Host Bus Device Support .......................................................................... 133
4.4.2.
Special Cycles ........................................................................................... 135
System Memory DRAM Interface ................................................................................ 136
4.5.1.
DRAM Organization and Configuration ..................................................... 136
4.5.1.1.
Configuration Mechanism For DIMMs .......................................... 137
4.5.1.2.
DRAM Register Programming ...................................................... 138
4.5.2.
DRAM Address Translation and Decoding................................................ 138
4.5.3.
DRAM Array Connectivity .......................................................................... 139
4.5.4.
SDRAMT Register Programming .............................................................. 140
4.5.5.
SDRAM Paging Policy............................................................................... 140
Intel Dynamic Video Memory Technology (D.V.M.T.)................................................ 140
Display Cache Interface............................................................................................... 141
4.7.1.
Supported DRAM Types for Display Cache Memory ................................ 141
4.7.2.
Memory Configurations ............................................................................. 142
4.7.3.
Address Translation .................................................................................. 143
4.7.4.
Display Cache Interface Timing ................................................................ 143
Internal Graphics Device ............................................................................................. 144
4.8.1.
3D/2D Instruction Processing.................................................................... 144
4.8.2.
3D Engine.................................................................................................. 145
4.8.3.
Buffers ....................................................................................................... 145
4.8.4.
Setup ......................................................................................................... 146
4.8.5.
Texturing.................................................................................................... 146
4.8.6.
2D Operation ............................................................................................. 148
4.8.7.
Fixed Blitter (BLT) and Stretch Blitter (STRBLT) Engines......................... 148
4.8.7.1.
Fixed BLT Engine ......................................................................... 149
4.8.7.2.
Arithmetic Stretch BLT Engine...................................................... 149
4.8.8.
Hardware Motion Compensation ............................................................... 149
4.8.9.
Hardware Cursor ....................................................................................... 150
4.8.10.
Overlay Engine .......................................................................................... 150
4.8.11.
Display....................................................................................................... 151
4.8.12.
Flat Panel/Digital CRT Interface / 1.8V TV-Out Interface.......................... 152
4.8.13.
DDC (Display Data Channel)..................................................................... 153
System Reset for the GMCH ....................................................................................... 154
System Clock Description............................................................................................ 154
Power Management .................................................................................................... 154
4.11.1.
Specifications Supported........................................................................... 154
Datasheet
82815 GMCH
R
5.
Pinout and Package Information ..............................................................................................155
5.1.
5.2.
6.
Datasheet
82815 GMCH Pinout ....................................................................................................155
Package Information ....................................................................................................162
Testability..................................................................................................................................165
6.1.
XOR TREE Testability Algorithm Example ..................................................................166
6.1.1.
Test Pattern Consideration for XOR Chains 3 and 4, and 7 and 8............166
6.2.
XOR Tree Initialization .................................................................................................167
6.2.1.
Chain [1:6] Initialization ..............................................................................167
6.2.2.
Chain [7:8] Initialization ..............................................................................167
6.3.
XOR Chain ...................................................................................................................168
6.4.
All Z ..............................................................................................................................172
7
82815 GMCH
R
Figures
Figure 1. Intel 815 Chipset Family System Block Diagram .................................................... 15
Figure 2. 82815 GMCH Block Diagram.................................................................................... 16
Figure 3. PAM Registers .......................................................................................................... 59
Figure 4. System Memory Address Map ................................................................................ 122
Figure 5. Detailed Memory System Address Map.................................................................. 122
Figure 6. DRAM Array Sockets .............................................................................................. 139
Figure 7. GMCH Display Cache Interface to 4 MB................................................................. 142
Figure 8. 3D/2D Pipeline Preprocessor.................................................................................. 144
Figure 9. Data Flow for the 3D Pipeline ................................................................................. 145
Figure 10. GMCH Pinout (Top View-Left Side) ...................................................................... 156
Figure 11. GMCH Pinout (Top View-Right Side).................................................................... 157
Figure 12. GMCH BGA Package Dimensions (Top and Side Views) .................................... 162
Figure 13. GMCH BGA Package Dimensions (Bottom View) ................................................ 163
Figure 14. XOR Tree Implementation .................................................................................... 165
Tables
Table 1. Supported System Bus and System Memory Bus Frequencies ................................ 20
Table 2. GMCH PCI Configuration Space (Device 0) .............................................................. 42
Table 3. Supported System Memory DIMM Configurations ..................................................... 54
Table 4. Attribute Bit Assignments ........................................................................................... 58
Table 5. PAM Registers and Associated Memory Segments .................................................. 59
Table 6. Summary of GMCH Error Sources, Enables and Status Flags ................................. 83
Table 7. GMCH Configuration Space (Device 1) ..................................................................... 84
Table 8. Device 2 Configuration Space Address Map (Internal Graphics)............................. 102
Table 9. Memory Segments and Their Attributes................................................................... 123
Table 10. Summay of Transactions Supported By GMCH..................................................... 133
Table 11. Host Responses Supported by the GMCH ............................................................ 134
Table 12. Special Cycles........................................................................................................ 135
Table 13. Sample Of Possible Mix And Match Options For 4 Row/2 DIMM
Configurations ........................................................................................................ 137
Table 14. Data Bytes on DIMM Used for Programming DRAM Registers ............................. 138
Table 15. GMCH DRAM Address Mux Function.................................................................... 139
Table 16. Programmable SDRAM Timing Parameters .......................................................... 140
Table 17. Memory Size for each configuration :..................................................................... 142
Table 18. GMCH Local Memory Address Mapping................................................................ 143
Table 19. Partial List of Display Modes Supported ................................................................ 151
Table 20. Partial List of Flat Panel Modes Supported ............................................................ 152
Table 21. Partial List of TV-Out Modes Supported ................................................................ 153
Table 22. Alphabetical Pin Assignment.................................................................................. 158
Table 23. Package Dimensions ............................................................................................. 163
Table 24. XOR Test Pattern Example.................................................................................... 166
Table 25 XOR Chain 1 35 Inputs Output: SMAA5 (A12) ...................................................... 168
Table 26 XOR Chain 2 33 Inputs Output: SMAA2 (F12) ...................................................... 168
Table 27 XOR Chain 3 38 Inputs Output: SMAA0 (D13) ...................................................... 169
Table 28 XOR Chain 4 36 Inputs Output: SMAA9 (D13) ...................................................... 169
Table 29 XOR Chain 5 56 Inputs Output: SMD31 (K5) ........................................................ 170
Table 30 XOR Chain 6 60 Inputs Output: SMAA11 (A13) .................................................... 171
Table 31 XOR Chain 7 33 Inputs Output: SMAA8 (D12) ...................................................... 171
Table 32 XOR Chain 8 31 Inputs Output: SMAA4 (B12) ...................................................... 172
8
Datasheet
82815 GMCH
R
Revision History
Rev.
-001
Datasheet
Description
Initial Release
Date
June 2000
9
82815 GMCH
R
This page is intentionally left blank.
10
Datasheet
82815 GMCH
R
82815 GMCH Features
Bus Support
! Processor/Host


 Intel Pentium III processor and Intel® Celeron™
Processor in FC-PGA package
 Supports processor 370-Pin Socket
 Supports 32-Bit System Bus Addressing
 4 deep in-order queue; 4 or 1 deep request queue
 Supports Uni-processor systems only
 In-order and Dynamic Deferred Transaction Support
 66/100/133MHz System Bus Frequency
 GTL+ I/O Buffer
! Integrated SDRAM Controller
 32 MB to 512 MB using 16Mb/64Mb/128Mb/256Mb
technology
 Supports up to 3 double sided DIMMs at 100 MHz system
memory bus
 Supports up to 2 double sided or 3 single sided DIMMs at
133 MHz system memory bus.
 64-bit data interface
 100/133 MHz system memory bus frequency
 Support for Asymmetrical SDRAM addressing only
 Support for x8 and x16 SDRAM device width
 Unbuffered, Non-ECC SDRAM only supported
 Refresh Mechanism: CBR ONLY supported
 Enhanced Open page arbitration SDRAM paging scheme
 Suspend to RAM support
! Accelerated Graphics Port (AGP) Interface Multiplexed with
Internal Graphics
 Supports a single AGP device via a connector
 Supports AGP 2.0 including 4x AGP data transfers
 AGP Universal Connector support via dual mode buffers
to allow AGP 2.0 3.3V or 1.5V signaling
 AGP PIPE# or SBA initiated accesses to SDRAM not
snooped
 AGP FRAME# initiated accesses to SDRAM are snooped
 High priority access support
 Hierarchical PCI configuration mechanism
 Delayed transaction support for AGP-to-SDRAM reads
that can not be serviced immediately
! Arbitration Scheme and Concurrency
 Intelligent Centralized Arbitration Model for Optimum
Concurrency Support
 Concurrent operations of processor and System busses
supported via dedicated arbitration and data buffering
! Data Buffering
 Distributed Data Buffering Model for optimum
concurrency
 SDRAM Write Buffer with read-around-write capability
 Dedicated processor –SDRAM, hub interface-SDRAM
and Graphics-SDRAM Read Buffers
! Power Management Functions
 SMRAM space remapping to A0000h (128 KB)
 Optional Extended SMRAM space above 256 MB,
additional 512 KB / 1MB TSEG from Top of Memory,
cacheable
 Stop Clock Grant and Halt special cycle translation from
the host to the hub interface
 ACPI Compliant power management
 APIC Buffer Management
 SMI, SCI, and SERR error indication
Datasheet
! Integrated Graphics Controller Multiplexed with AGP
Controller
 3D Hyper Pipelined Architecture
 -Parallel Data Processing (PDP)
 -Precise Pixel Interpolation (PPI)
 Full 2D H/W Acceleration
 Motion Video Acceleration
 Supports 133 MHz System Memory while running in
non-CPC mode
! 3D Graphics Visual Enhancements







Flat & Gouraud Shading
Mip Maps with Trilinear and Anisotropic Filtering
Full Color Specular
Fogging Atmospheric Effects
Z Buffering
3D Pipe 2D Clipping
Backface Culling
! 3D Graphics Texturing Enhancements
 Per Pixel Perspective Correction Texture Mapping
 Texture Compositing
 Texture Color Keying/Chroma Keying
! Digital Video Output
 85 MHz Flat Panel Monitor/Digital CRT Interface Or
Digital Video Output for use with a external TV encoder
! Display
 Integrated 24-bit 230 MHz RAMDAC
 Gamma Corrected Video
 DDC2B Compliant
! 2D Graphics
 Up to 1600x1200 in 8-bit Color at 85 Hz Refresh
 Hardware Accelerated Functions
 3 Operand Raster BitBLTs
 64x64x3 Color Transparent Cursor
! Arithmetic Stretch Blitter Video
 H/W Motion Compensation Assistance for S/W MPEG2
Decode
 Software DVD at 30 fps
 Digital Video Out Port
 NTSC and PAL TV Out Support
 H/W Overlay Engine with Bilinear Filtering
 Independent gamma correction, saturation, brightness &
contrast for overlay
! Integrated Graphics Memory Controller
 Intel D.V.M. Technology
! Display Cache Interface multiplexed on the AGP interface
 32-bit data interface
 133 MHz SDRAM interface only.
 Flexible AGP In-Line Memory Module (AIMM)
Implementation
 Support for 2 1Mx16, or 1 2Mx32 on AIMM card
 4 MB maximum addressable
! Supporting I/O Bridge
 82801AA I/O Controller Hub (ICH)
 82801BA I/O Controller Hub (ICH2)
! Packaging/Power
 544 BGA
 1.85V core with 3.3V CMOS I/O
11
82815 GMCH
R
82815 GMCH Simplified Block Diagram
HA[31:3]#
HD[63:0]#
ADS#
BNR#
BPRI#
DBSY#
DEFER#
DRDY#
HIT#
HITM#
HLOCK#
HREQ[4:0]#
HTRDY#
RS[2:0]#
CPURST#
GTLREF[1:0]
SMAA[12:0]
SMAB[7:4]#
SMAC[7:4]#
SBS[1:0]
SMD[63:0]
SDQM[7:0]
SCSA[5:0]#
SCSB[5:0]#
SRAS#
SCAS#
SWE#
SCKE[5:0]
SRCOMP
HCLK
SCLK
LTCLK[1:0]
LOCLK
LRCLK
DCLKREF
HLCLK
RESET#
HLREF
HL[10:0]
HLSTRB
HLSTRB#
HLZCOMP
12
Display
Interface
Host Bus
Interface
Digital
Video
Out
System
Memory
Interface
Clock
And Reset
Signals
Hub
Interface
Display
Cache
Interface
AGP
Interface
VSYNC
HSYNC
IREF
RED
GREEN
BLUE
DDCK
DDDA
LTVCL
LTVDA
TVCLKIN/INT#
LTVCLKOUT[1:0]
LTVBLANK#
LTVDATA[11:0]
LTVSYNC
LTVHSYNC
LCS#
LDQM[3:0]
LSRAS#
LSCAS#
LMA[11:0]
LWE#
LMD[31:0]
PIPE#
SBA[7:0]
RBF#
WBF#
ST[2:0]
AD_STB[1:0]
AD_STB[1:0]#
SB_STB
SB_STB#
G_FRAME#
G_IRDY#
G_TRDY#
G_STOP#
G_DEVSEL#
G_REQ#
G_GNT#
G_AD[31:0]#
G_C/BE[3:0]#
G_PAR
GRCOMP
AGPREF
Datasheet
82815 GMCH
R
1.
Overview
The Intel 815 chipset family is a high-flexibility chipset designed to extend from the basic
graphics/multimedia PC platform up to the mainstream performance desktop platform. The chipset
consists of a 82815 Graphics and Memory Controller Hub (GMCH) and an I/O Controller Hub for the
I/O subsystem. The GMCH integrates a system memory SDRAM controller that supports a
64-bit 100/133 MHz SDRAM array.
The 82815 GMCH integrates a Display Cache SDRAM controller that supports a 32-bit 133 MHz
SDRAM array for enhanced integrated 2D and 3D graphics performance. Multiplexed with the display
cache interface is an AGP controller interface to enable graphics configuration and upgrade flexibility
with the Intel 815 chipset family. The AGP interface and the internal graphics device are mutually
exclusive. When the AGP port is populated with an AGP graphics card, the integrated graphics is
disabled; thus, the display cache interface is not needed.
There are two chipsets in the Intel® 815 chipset family:
• Intel® 815 chipset: This chipset contains the 82815 GMCH and 82801AA ICH.
• Intel® 815E chipset. This chipset contains the 82815 GMCH and 82801BA ICH2.
Note:
The only component difference between the Intel® 815 chipset and the Intel® 815E chipset is the I/O
Controller Hub.
This datasheet provides an overview of the 815 chipset family (see Section 1.2). The remainder of the
document describes the Intel 82815 Graphics Memory Controller Hub (GMCH).
The Intel 815 chipset family may contain design defects or errors known as errata which may cause the product to deviate from
published specifications. Current characterized errata are available on request.
1.1.
Related Documents
• AGTL+ I/O Specification: Contained in the Intel Pentium II Processor Databook
• PCI Local bus Specification 2.2: Contact www.pcisig.com
• Intel® 82801AA (ICH) and Intel® 82801AB (ICH0) I/O Controller Hub Datasheet
(Document Number 290655)
• Intel® 82801BA I/O Controller Hub (ICH2) Datasheet (Document Number 290687)
• Intel® 82802 Firmware Hub (FWH) Datasheet (Document Number 290658)
• Intel® 815 Chipset Design Guide (Document Number 298233)
• Intel® 815E Chipset Design Guide (Document Number 298234)
Datasheet
13
82815 GMCH
R
1.2.
The Intel 815 Chipset Family
Figure 1 shows a typical system block diagram based on the Intel® 815 chipset family. The chipset uses a
hub architecture with the GMCH as the host bridge hub and the I/O Controller Hub as the I/O hub. The
GMCH supports processor bus frequencies of 66/100/133 MHz. The I/O Controller Hub is highly
integrated providing many of the functions needed in today’s PC platforms; it also provides the interface
to the PCI Bus. The GMCH and I/O Controller Hub communicate over a dedicated hub interface.
82801AA ICH and 82801BA ICH2 functions include:
• PCI Rev 2.2 compliant with support for 33 MHz PCI operations
• Supports up to 6 Req/Gnt pairs (PCI Slots)
• Power management logic support
• Enhanced DMA controller, interrupt controller, and timer functions
• Integrated IDE controller
 Ultra ATA/66/33 (ICH)
 Ultra ATA/100/66/33 (ICH2)
• USB host interface
 1 host controller and supports 2 USB ports (ICH)
 2 host controllers and supports 4 USB ports (ICH2)
• Integrated LAN controller (ICH2 only)
• System Management Bus (SMBus) compatible with most I2C devices
 ICH has bus master capability
 ICH2 has both bus master and slave capability
• AC’97 2.1 compliant link for audio and telephony codecs
 2 channels (ICH)
 Up to 6 channels (ICH2)
• Low Pin Count (LPC) interface
• Firmware Hub (FWH) interface support
 Intel’s FWH component is the 82802: It contains a Random Number Generator (RNG), five
General Purpose Inputs (GPIs), register-based block locking, hardware-based locking, and
Flash memory for platform code/data nonvolatile storage
 FWH component is also available from other suppliers
• Alert on LAN*
 AOL (ICH and ICH2)
 AOL2 (ICH2 only)
14
Datasheet
82815 GMCH
R
Figure 1. Intel 815 Chipset Family System Block Diagram
®
Intel® Pentium
III Processor
or
®
Intel Celeron™ Processor
System Bus (66/100/133 MHz)
Digital Video Out
Intel® 815 Chipset family
Encoder
Digital Video Out
64 Bit /
82815 GMCH
100/133 MHz Only System
(Graphics and Memory
Memory
Controller Hub)
TV
AGP Connector
AGP Graphics
Analog Display
Or
AGP 2.0
Display Cache
(4 MB SDRAM,
133 MHz Only)
- Memory Controller
- AGP Contoller
- Graphcs Controller
- 3D Engine
- 2D Engine
- Video Engine
Hub
Interface
4 IDE Drives
UltraATA/66/33 (ICH)
UltraATA/100/66/33 (ICH2)
PCI
Slots
2 USB Ports; 1 HC (ICH)
4 USB Ports; 2 HC (ICH2)
PCI Bus
AC'97 Codec(s) AC'97 2.1
(optional)
(ICH and
ICH2)
Keyboard,
Mouse, FD, PP,
SP, IR
I/O Controller Hub
(82801AA ICH
and
82801BA ICH2)
(ICH and
ICH2)
ISA
Bridge
(optional)
ISA
Slots
LPC I/F
Super I/O
PCI
Agent
(ICH and
ICH2)
LAN Connect
GPIO
(ICH2
only)
(ICH and
ICH2)
FWH
815_SysBlk
Datasheet
15
82815 GMCH
R
1.3.
82815 GMCH Overview
Figure 2 is a block diagram of the GMCH illustrating the various interfaces and integrated functions. The
functions and capabilities include:
• Support for a single processor configuration
• 64-bit AGTL+ based System Bus Interface at 66/100/133 MHz
• 32-bit Host Address Support
• 64-bit System Memory Interface with optimized support for SDRAM at 100/133 MHz
• Integrated 2D & 3D Graphics Engines
• Integrated H/W Motion Compensation Engine
• Integrated 230 MHz DAC
• Integrated Digital Video Out Port
• 133 MHz Display Cache
• AGP 1X/2X/4X Controller
Figure 2. 82815 GMCH Block Diagram
System Bus Interface
Analog
Display
Out
Display Engine
3D Engine
HW Motion Comp
3D
Engine
DAC
Overlay
2D Engine
HW Cursor
Digital
Video
Out
Buffer
Digital Video Out
Port
Stretch
BLT Eng
Memory Interface
System
Memory
BLT Eng
DDC/
I2C
AGP Interface
AGP/
Display
Cache
Pins
Local Memory
Interface
Buffer
Hub Interface
gmch_blk2.vsd
16
Datasheet
82815 GMCH
R
1.4.
Host Interface
The host interface of the GMCH is optimized to support the Intel Pentium® III processor and Intel
CeleronTM processor in the FC-PGA package. The GMCH implements the host address, control, and data
bus interfaces within a single device. The GMCH supports a 4-deep in-order queue (i.e., supports
pipelining of up to 4 outstanding transaction requests on the host bus) . Host bus addresses are decoded
by the GMCH for accesses to system memory, PCI memory and PCI I/O (via hub interface), PCI
configuration space and Graphics memory. The GMCH takes advantage of the pipelined addressing
capability of the processor to improve the overall system performance.
The 82815 GMCH supports the 370-pin socket processor.
• 370-pin socket (PGA370). The PGA370 is a zero insertion force (ZIF) socket that a processor in
the FC-PGA package will use to interface with a system board.
1.5.
System Memory Interface
The GMCH integrates a system memory controller that supports a 64-bit 100/133 MHz SDRAM array.
The only DRAM type supported is industry standard Synchronous DRAM (SDRAM). The SDRAM
controller interface is fully configurable through a set of control registers.
The GMCH supports industry standard 64-bit wide DIMMs with SDRAM devices. The thirteen
multiplexed address lines (SMAA[12:0]) along with the two bank select lines (SBS[1:0]) allow the
GMCH to support 2M, 4M, 8M, 16M, and 32M x64 DIMMs. Only asymmetric addressing is supported.
The GMCH has 6 SCS# lines (2 copies of each for electrical loading), enabling the support of up to six
64-bit rows of SDRAM. The GMCH targets SDRAM with CL2 and CL3, and supports both single and
double-sided DIMMs. Additionally, the GMCH also provides a 1024 deep refresh queue. The GMCH
can be configured to keep up to 4 pages open within the memory array. Pages can be kept open in any
one bank of memory.
The Intel 815 chipset family supports up to 3 DIMM connectors in a system. A maximum of 2 doublesided or 3 single-sided DIMMs may be populated when the SDRAM interface is operating at 133 MHz.
Upon detection that additional rows are populated beyond these configurations, the BIOS must downshift the SDRAM clocks to 100 MHz through a two-wire interface of the system clock generator.
SCKE[5:0] is used in configurations requiring powerdown mode for the SDRAM.
Datasheet
17
82815 GMCH
R
1.6.
Multiplexed AGP and Display Cache Interface
The 82815 GMCH multiplexes an AGP interface with a display cache interface for internal 3D graphics
performance improvement. The display cache is used only in the internal graphics. When an AGP card is
installed in the system, the GMCH internal graphics will be disabled and the AGP controller will be
enabled.
1.6.1.
AGP Interface
A single AGP connector is supported by the GMCH AGP interface. The AGP buffers operate in one of
two selectable modes in order to support the AGP Universal Connector:
• 3.3V drive, not 5 volt safe: This mode is compliant to the AGP 1.0 and 2.0 specifications.
• 1.5V drive, not 3.3 volt safe: This mode is compliant with the AGP 2.0 specification.
The following table shows the AGP Data Rate and the Signaling Levels supported by the GMCH.
Data Rate
Signaling Level
1.5V
3.3V
1x AGP
Yes
Yes
2x AGP
Yes
Yes
4x AGP
Yes
No
The AGP interface supports 4x AGP signaling. AGP semantic (PIPE# or SBA[7:0]) cycles to SDRAM
are not snooped on the host bus. AGP FRAME# cycles to SDRAM are snooped on the host bus. The
GMCH supports PIPE# or SBA[7:0] AGP address mechanisms, but not both simultaneously. Either the
PIPE# or the SBA[7:0] mechanism must be selected during system initialization. High priority accesses
are supported. Only memory writes from the hub interface to AGP are allowed. No transactions from
AGP to the hub interface are allowed.
1.6.2.
Display Cache Interface
The GMCH supports a Display Cache SDRAM controller with a 32-bit 133 MHz SDRAM array. The
DRAM type supported is industry standard Synchronous DRAM (SDRAM) like that of the system
memory. The local memory SDRAM controller interface is fully configurable through a set of control
registers.
1.7.
Hub Interface
The hub interface is a private interconnect between the GMCH and the I/O Controller Hub.
18
Datasheet
82815 GMCH
R
1.8.
82815 GMCH Integrated Graphics Support
The GMCH includes a highly integrated graphics accelerator. Its architecture consists of dedicated multimedia engines executing in parallel to deliver high performance 3D, 2D, and motion compensation video
capabilities. The 3D and 2D engines are managed by a 3D/2D pipeline preprocessor allowing a sustained
flow of graphics data to be rendered and displayed. The deeply pipelined 3D accelerator engine provides
3D graphics quality and performance via per-pixel 3D rendering and parallel data paths which allow each
pipeline stage to simultaneously operate on different primitives or portions of the same primitive. The
GMCH graphics accelerator engine supports perspective-correct texture mapping, trilinear and
anisotropic Mip-Map filtering, Gouraud shading, alpha-blending, fogging and Z-buffering. A rich set of
3D instructions permit these features to be independently enabled or disabled.
For the GMCH, a Display Cache (DC) can be used for the Z-buffer (textures and display buffer(s) are
located only in system memory). If the display cache is not used, the Z-buffer is located in system
memory.
The GMCH integrated graphics accelerator’s 2D capabilities include BLT and arithmetic STRBLT
engines, a hardware cursor and an extensive set of 2D registers and instructions. The high performance
64-bit BitBLT engine provides hardware acceleration for many common Windows operations.
In addition to its 2D/3D capabilities, the GMCH integrated graphics accelerator also supports full
MPEG-2 motion compensation for software-assisted DVD video playback, a VESA DDC2B compliant
display interface and a digital video out port which may support (via an external video encoder) NTSC
and PAL broadcast standards and (via an external TMDS transmitter) digital Flat Panel or Digital CRT
displays.
1.8.1.
Display, Digital Video Out, and LCD/Flat Panel/Digital CRT
The GMCH provides interfaces to a standard progressive scan monitor, TV-Out device, and TMDS
transmitter. These interfaces are only active when running in internal graphics mode.
• The GMCH directly drives a standard progressive scan monitor up to a resolution of 1600x1200
pixels.
• The GMCH provides a Digital Video Out interface to connect an external device to drive a
1280x1024 resolution non-scalar DDP digital Flat Panel with appropriate EDID 1.2 data or digital
CRTs. The interface has 1.8V signaling to allow it to operate at higher frequencies. This interface
can also connect to a 1.8V TV-Out encoder.
Datasheet
19
82815 GMCH
R
1.9.
System Clocking
The 82815 GMCH has a new type of clocking architecture. It has integrated SDRAM buffers that run at
either 100 or 133 MHz, independent of the system bus frequency. See table below for supported system
bus and system memory bus frequencies. The system bus frequency is selectable between 66 MHz,
100 MHz, or 133 MHz. The GMCH uses a copy of the USB clock as the DOT Clock input for the
graphics pixel clock PLL.
Table 1. Supported System Bus and System Memory Bus Frequencies
1.10.
Front Side Bus
Frequency
System Memory
Bus Frequency
Display Cache Interface
Frequency
66 MHz
100 MHz
133 MHz or DVMT
100 MHz
100 MHz
133 MHz or DVMT
133 MHz
100 MHz
133 MHz or DVMT
133 MHz
133 MHz
133 MHz or DVMT
GMCH Power Delivery
The 82815 GMCH core voltage is 1.85V. System memory operates from a 3.3V supply. Display cache
memory operates from the AGP 3.3V supply. AGP 1X/2X I/O can operate from either a 3.3V or a 1.5V
supply. AGP 4X I/O requires a 1.5V supply. The AGP interface voltage is determined by the VDDQ
generation on the motherboard.
20
Datasheet
82815 GMCH
R
2.
Signal Description
This section provides a detailed description of the GMCH signals. The signals are arranged in functional
groups according to their associated interface. The states of all of the signals during reset are provided in
the System Reset section.
The “#” symbol at the end of a signal name indicates that the active, or asserted state occurs when the
signal is at a low voltage level. When “#” is not present after the signal name the signal is asserted when
at the high voltage level.
The following notations are used to describe the signal type:
I
Input pin
O
Output pin
I/OD
Input / Open Drain Output pin. This pin requires a pullup
I/O
Bi-directional Input/Output pin
s/t/s
Sustained Tristate. This pin is driven to its inactive state prior to tri-stating.
As/t/s
Active Sustained Tristate. This applies to some of the hub interface signals. This pin
is weakly driven to its last driven value.
The signal description also includes the type of buffer used for the particular signal:
AGTL+
Open Drain AGTL+ interface signal. Refer to the AGTL+ I/O Specification for
complete details
AGP
AGP interface signals. These signals can be programmed to be compatible with AGP
2.0 3.3V or 1.5V Signaling Environment DC and AC Specifications. In 3.3V mode
the buffers are not 5V tolerant. In 1.5V mode the buffers are not 3.3V tolerant.
CMOS
The CMOS buffers are low voltage TTL compatible signals. These are 3.3V only.
LVTTL
Low Voltage TTL compatible signals. There are 3.3V only.
1.8V
1.8V signals for the digital video interface
Analog
Analog CRT Signals
Note that the processor address and data bus signals (Host Interface) are logically inverted signals
(i.e., the actual values are inverted of what appears on the processor bus). This must be taken into
account and the addresses and data bus signals must be inverted inside the GMCH. All processor control
signals follow normal convention. A 0 indicates an active level (low voltage) if the signal is followed by
a # symbol and a 1 indicates an active level (high voltage) if the signal has no # suffix.
Datasheet
21
82815 GMCH
R
2.1.
Host Interface Signals
Signal Name
Type
Description
CPURST#
O
AGTL+
CPU Reset. The GMCH asserts CPURST# while RESET# (PCIRST# from the I/O
Controller Hub) is asserted and for approximately 1 ms after RESET# is deasserted.
The GMCH also pulses CPURST# for approximately 1 ms when requested via a
hub interface special cycle. The CPURST# allows the processor to begin execution
in a known state.
HA[31:3]#
I/O
AGTL+
Host Address Bus. HA[31:3]# connect to the processor address bus. During
processor cycles, HA[31:3]# are inputs. The GMCH drives HA[31:3]# during snoop
cycles on behalf of Primary PCI. Note that the address bus is inverted on the
processor bus.
HD[63:0]#
I/O
AGTL+
Host Data. These signals are connected to the processor data bus. Note that the
data signals are inverted on the processor bus.
ADS#
I/O
AGTL+
Address Strobe. The processor bus owner asserts ADS# to indicate the first of two
cycles of a request phase.
BNR#
I/O
AGTL+
Block Next Request. Used to block the current request bus owner from issuing a
new request. This signal is used to dynamically control the processor bus pipeline
depth.
BPRI#
O
AGTL+
Priority Agent Bus Request. The GMCH is the only priority agent on the processor
bus. It asserts this signal to obtain the ownership of the address bus. This signal
has priority over symmetric bus requests and will cause the current symmetric
owner to stop issuing new transactions unless the HLOCK# signal was asserted.
DBSY#
I/O
AGTL+
Data Bus Busy. Used by the data bus owner to hold the data bus for transfers
requiring more than one cycle.
DEFER#
O
AGTL+
Defer. The GMCH will generate a deferred response as defined by the rules of the
GMCH dynamic defer policy. The GMCH will also use the DEFER# signal to
indicate a processor retry response.
DRDY#
I/O
AGTL+
Data Ready. Asserted for each cycle that data is transferred.
HIT#
I/O
AGTL+
Hit. Indicates that a caching agent holds an unmodified version of the requested
line. Also driven in conjunction with HITM# by the target to extend the snoop
window.
HITM#
I/O
AGTL+
Hit Modified. Indicates that a caching agent holds a modified version of the
requested line and that this agent assumes responsibility for providing the line.
HITM# is also driven in conjunction with HIT# to extend the snoop window.
HLOCK#
I
AGTL+
Host Lock. All processor bus cycles sampled with the assertion of HLOCK# and
ADS#, until the negation of HLOCK# must be atomic (i.e., no hub interface or
GMCH graphics snoopable access to SDRAM is allowed when HLOCK# is asserted
by the processor).
HREQ[4:0]#
I/O
AGTL+
Host Request Command. Asserted during both clocks of request phase. In the first
clock, the signals define the transaction type to a level of detail that is sufficient to
begin a snoop request. In the second clock, the signals carry additional information
to define the complete transaction type.
The transactions supported by the GMCH are defined in the Host Interface section
of this document.
HTRDY#
22
I/O
AGTL+
Host Target Ready. Indicates that the target of the processor transaction is able to
enter the data transfer phase.
Datasheet
82815 GMCH
R
Signal Name
Type
RS[2:0]#
I/O
AGTL+
Description
Response Signals. Indicates type of response as shown below:
000 = Idle state
001 = Retry response
010 = Deferred response
011 = Reserved (not driven by the GMCH)
100 = Hard Failure (not driven by the GMCH)
101 = No data response
110 = Implicit Writeback
111 = Normal data response
GTLREF[1:0]
2.2.
I
GTL Reference. Reference voltage input for the Host GTL interface. GTLREF is
2/3 * VTT. VTT is nominally 1.5V.
System Memory Interface Signals
Signal Name
Type
Description
SMAA[12:0]
SMAB[7:4]#
SMAC[7:4]#
SBS[1:0]
O
CMOS
Memory Address. SMAA[12:0], SMAB[7:4]#, and SMAC[7:4]# are used to provide
the multiplexed row and column address to SDRAM. SBS[1:0] provide the Bank
Select.
SBS[1:0]
O
CMOS
Memory Bank Select. These signals define the banks that are selected within
each DRAM row. The SMAx and SBS signals combine to address every possible
location within a DRAM device.
SBS[1:0] may be heavily loaded and require 2 SDRAM clock cycles for setup time
to the SDRAMs. For this reason, all chip select signals (SCSA[5:0]# and
SCSB[5:0]#) must be deasserted on any SDRAM clock cycle that one of these
signals change.
Datasheet
SMD[63:0]
I/O
CMOS
Memory Data. These signals are used to interface to the SDRAM data bus.
SDQM[7:0]
O
CMOS
Input/Output Data Mask. These pins act as synchronized output enables during
read cycles and as a byte enables during write cycles.
SCSA[5:0]#
SCSB[5:0]#
O
CMOS
Chip Select. For the memory row configured with SDRAM, these pins perform the
function of selecting the particular SDRAM components during the active state.
SRAS#
O
CMOS
SDRAM Row Address Strobe. These signals drive the SDRAM array directly
without any external buffers.
SCAS#
O
CMOS
SDRAM Column Address Strobe. These signals drive the SDRAM array directly
without any external buffers.
SWE#
O
CMOS
Write Enable Signal. SWE# is asserted during writes to SDRAM.
SCKE[5:0]
O
CMOS
System Memory Clock Enable. SCKE SDRAM Clock Enable is used to signal a
self-refresh or power-down command to an SDRAM array when entering system
suspend.
SRCOMP
O
System Memory RCOMP. Used to calibrate the System memory I/O buffers. This
pin should be connected to a 40 ohm resistor tied to 3.3V VCC (VSUS3.3).
23
82815 GMCH
R
2.3.
AGP Interface Signals
For more details on the operation of these signals, refer to the AGP Interface Specification Revision 2.0.
Some of the AGP interface signals are multiplexed with Display Cache interface signals. AGP inteface
signals only function as documented in this section when the GMCH AGP interface is enabled (GMCH
integrated graphics disabled). Refer to Section 2.11 for multiplexing map of AGP to Display Cache
interface signals.
2.3.1.
AGP Addressing Signals
There are two mechanisms that the AGP master can enqueue AGP requests: PIPE# and SBA (side-band
addressing). Upon initialization, one of the methods is chosen. The master may not switch methods
without a full reset of the system. When PIPE# is used to enqueue addresses, the master is not allowed to
queue addresses using the SBA bus. For example, during configuration time, if the master indicates that
it can use either mechanism, the configuration software will indicate which mechanism the master will
use. Once this choice has been made, the master will continue to use the mechanism selected until the
system is reset (and reprogrammed) to use the other mode. This change of modes is not a dynamic
mechanism but rather a static decision when the device is first being configured after reset.
Signal Name
Type
PIPE#
I
AGP
Description
Pipeline.
During PIPE# Operation. This signal is asserted by the AGP master to indicate a
full-width address is to be enqueued on by the target using the AD bus. One
address is placed in the AGP request queue on each rising clock edge while PIPE#
is asserted.
During SBA Operation. This signal is not used if SBA (Side Band Addressing) is
selected.
During FRAME# Operation. This signal is not used during AGP FRAME#
operation.
SBA[7:0]
I
AGP
Side-band Addressing.
During PIPE# Operation. These signals are not used during PIPE# operation.
During SBA Operation. These signals (the SBA, or side-band addressing, bus) are
used by the AGP master (graphics component) to place addresses into the AGP
request queue. The SBA bus and AD bus operate independently. That is,
transactions can proceed on the SBA bus and the AD bus simultaneously.
During FRAME# Operation. These signals are not used during AGP FRAME#
operation.
24
Datasheet
82815 GMCH
R
2.3.2.
AGP Flow Control Signals
Signal Name
Type
RBF#
I
AGP
Description
Read Buffer Full.
During PIPE# and SBA Operation. Read buffer full indicates if the master is ready
to accept previously requested low priority read data. When RBF# is asserted the
GMCH is not allowed to initiate the return low priority read data. That is, the GMCH
can finish returning the data for the request currently being serviced, however it can
not begin returning data for the next request. RBF# is only sampled at the beginning
of a cycle.
If the AGP master is always ready to accept return read data, then it is not required
to implement this signal.
During FRAME# Operation. This signal is not used during AGP FRAME#
operation.
WBF#
I
AGP
Write-Buffer Full.
During PIPE# and SBA Operation. Write bufffer full indicates if the master is ready
to accept Fast Write data from the GMCH. When WBF# is asserted the GMCH is
not allowed to drive Fast Write data to the AGP master. WBF# is only sampled at
the beginning of a cycle.
If the AGP master is always ready to accept fast write data, then it is not required to
implement this signal.
During FRAME# Operation: This signal is not used during AGP FRAME#
operation.
2.3.3.
AGP Status Signals
Signal Name
Type
ST[2:0]
O
AGP
Description
Status Bus.
During PIPE# and SBA Operation. Provides information from the arbiter to a AGP
Master on what it may do. ST[2:0] only have meaning to the master when its GNT#
is asserted. When GNT# is deasserted, these signals have no meaning and must
be ignored. Refer to the AGP Interface Specificaiton revision 2.0 for further
explanation of the ST[2:0] values and their meanings.
During FRAME# Operation. These signals are not used during FRAME# based
operation; except that a ‘111’ indicates that the master may begin a FRAME#
transaction.
Datasheet
25
82815 GMCH
R
2.3.4.
AGP Clocking Signals (Strobes)
Signal Name
Type
AD_STB0
I/O
s/t/s
AGP
Description
AD Bus Strobe-0.
During 2X Operation. During 2X operation, this signal provides timing for the
G_AD[15:0] and G_C/BE[1:0]# signals. The agent that is providing the data will
drive this signal.
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the G_AD[15:0] and G_C/BE[1:0]# signals.
AD_STB0#
I/O
s/t/s
AGP
AD Bus Strobe-0 Compliment.
During 2X Operation. During 2X operation, this signal is not used.
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the G_AD[15:0] and G_C/BE[1:0]# signals.
The agent that is providing the data will drive this signal.
AD_STB1
I/O
s/t/s
AGP
AD Bus Strobe-1.
During 2X Operation. During 2X operation, this signal provides timing for the
G_AD[16:31] and G_C/BE[2:3]# signals. The agent that is providing the data drives
this signal.
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the G_AD[16:31] and G_C/BE[2:3]# signals.
The agent that is providing the data drives this signal.
AD_STB1#
I/O
s/t/s
AGP
AD Bus Strobe-1 Compliment.
During 2X Operation. During 2X operation, this signal is not used
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the G_AD[16:31] and G_C/BE[2:3]# signals.
The agent that is providing the data drives this signal.
SB_STB
I
AGP
SBA Bus Strobe.
During 2X Operation. During 2X operation, this signal provides timing for the SBA
bus signals. The agent that is driving the SBA bus drives this signal.
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the SBA bus signals. The agent that is
driving the SBA bus drives this signal.
SB_STB#
I
AGP
SBA Bus Strobe Compliment.
During 2X Operation. During 2X operation, this signal is not used.
During 4X Operation. During 4X operation, this is one-half of a differential strobe
pair that provides timing information for the SBA bus signals. The agent that is
driving the SBA bus drives this signal.
26
GRCOMP
O
AGP RCOMP. Used to calibrate AGP I/O buffers. This pin should be connected to a
40 ohm pull down resistor tied to VSS.
AGPREF
I
AGP Reference. Reference voltage input for the AGP interface. AGPREF should
be 0.4*VDDAGP when VDD is 3.3V, or 0.5* VDDAGP when VDD is 1.5V.
Datasheet
82815 GMCH
R
2.3.5.
AGP FRAME# Signals
Signal Name
Type
G_FRAME#
I/O
s/t/s
AGP
Description
FRAME.
During PIPE# and SBA Operation. Not used by AGP SBA and PIPE#, but used
during AGP FRAME# .
During Fast Write Operation. G_FRAME# is used to frame transactions as an
output from the GMCH during Fast Writes.
During FRAME# Operation. G_FRAME# is an output when the GMCH acts as an
initiator on the AGP Interface. G_FRAME# is asserted by the GMCH to indicate the
beginning and duration of an access. G_FRAME# is an input when the GMCH acts
as a FRAME# based AGP target. As a FRAME# based AGP target, the GMCH
latches the G-C/BE[3:0]# and the G_AD[31:0] signals on the first clock edge on
which it samples G_FRAME# active.
G_IRDY#
I/O
s/t/s
AGP
Initiator Ready.
During PIPE# and SBA Operation. Not used while enqueueing requests via AGP
SBA and PIPE#, but used during the data phase of PIPE# and SBA transactions.
During FRAME# Operation. G_IRDY# is an output when GMCH acts as a
FRAME# based AGP initiator and an input when the GMCH acts as a FRAME#
based AGP target. The assertion of G_IRDY# indicates the current FRAME# based
AGP bus initiator’s ability to complete the current data phase of the transaction.
During Fast Write Operation. G_IRDY# indicates the AGP compliant master is
ready to provide all write data for the current transaction. Once G_IRDY# is
asserted for a write operation, the master is not allowed to insert wait states. The
master is never allowed to insert a wait state during the initial data transfer (32
bytes) of a write transaction. However, it may insert wait states after each 32 byte
block is transferred.
G_TRDY#
I/O
s/t/s
AGP
Target Ready.
During PIPE# and SBA Operation. Not used while enqueueing requests via AGP
SBA and PIPE#, but used during the data phase of PIPE# and SBA transactions.
During FRAME# Operation. G_TRDY# is an input when the GMCH acts as an
AGP initiator and an output when the GMCH acts as a FRAME# based AGP target.
The assertion of G_TRDY# indicates the target’s ability to complete the current data
phase of the transaction.
During Fast Write Operation. G_TRDY# indicates the AGP compliant target is
ready to receive write data for the entire transaction (when the transfer size is less
than or equal to 32 bytes) or is ready to transfer the initial or subsequent block (32
bytes) of data when the transfer size is greater than 32 bytes. The target is allowed
to insert wait states after each block (32 bytes) is transferred on write transactions.
G_STOP#
I/O
s/t/s
AGP
Stop.
During PIPE# and SBA Operation. This signal is not used during PIPE# or SBA
operation.
During FRAME# Operation. STOP# is an input when the GMCH acts as a
FRAME# based AGP initiator and an output when the GMCH acts as a FRAME#
based AGP target. STOP# is used for disconnect, retry, and abort sequences on
the AGP interface.
Datasheet
27
82815 GMCH
R
Signal Name
Type
G_DEVSEL#
I/O
s/t/s
AGP
Description
Device Select.
During PIPE# and SBA Operation. This signal is not used during PIPE# or SBA
operation.
During FRAME# Operation. G_DEVSEL#, when asserted, indicates that a
FRAME# based AGP target device has decoded its address as the target of the
current access. The GMCH asserts G_DEVSEL# based on the SDRAM address
range being accessed by a PCI initiator. As an input it indicates whether any device
on the bus has been selected.
G_REQ#
I
AGP
Request.
During SBA Operation. This signal is not used during SBA operation.
During PIPE# and FRAME# Operation. G_REQ#, when asserted, indicates that a
FRAME# or PIPE# based AGP master is requesting use of the AGP interface. This
signal is an input into the GMCH.
G_GNT#
G_AD[31:0]
O
AGP
Grant.
I/O
AGP
Address/Data Bus.
During SBA, PIPE# and FRAME# Operation. G_GNT# along with the information
on the ST[2:0] signals (status bus) indicates how the AGP interface will be used
next. Refer to the AGP Interface Specificaiton revision 2.0 for further explanation of
the ST[2:0] values and their meanings.
During PIPE# and FRAME# Operation. G_AD[31:0] are used to transfer both
address and data information on the AGP inteface.
During SBA Operation. G_AD[31:0] are used to transfer data on the AGP
interface.
G_C/BE[3:0]#
I/O
AGP
Command/Byte Enable.
During FRAME# Operation. During the address phase of a transaction,
G_C/BE[3:0]# define the bus command. During the data phase G_C/BE[3:0]# are
used as byte enables. The byte enables determine which byte lanes carry
meaningful data. The commands issued on the G_C/BE# signals during FRAME#
based AGP are the same G_C/BE# command described in the PCI 2.1 and 2.2
specifications.
During PIPE# Operation. When an address is enqueued using PIPE#, the C/BE#
signals carry command information. Refer to the AGP 2.0 Interface Specification
Revision 2.0 for the definition of these commands. The command encoding used
during PIPE# based AGP is Different than the command encoding used during
FRAME# based AGP cycles (or standard PCI cycles on a PCI bus).
During SBA Operation. These signals are not used during SBA operation.
G_PAR
I/O
AGP
Parity.
During FRAME# Operation. G_PAR is driven by the GMCH when it acts as a
FRAME# based AGP initiator during address and data phases for a write cycle, and
during the address phase for a read cycle. G_PAR is driven by the GMCH when it
acts as a FRAME# based AGP target during each data phase of a FRAME# based
AGP memory read cycle. Even parity is generated across G_AD[31:0] and
G_C/BE[3:0]#.
During SBA and PIPE# Operation. This signal is not used during SBA and PIPE#
operation.
NOTES:
1. LOCK#, SERR#, and PERR# signals are not supported on the AGP Interface (even for PCI operations).
2. PCI signals described in this table behave according to PCI 2.1 specifications when used to perform PCI
transactions on the AGP interface.
28
Datasheet
82815 GMCH
R
2.4.
Display Cache Interface Signals
Some of the Display Cache interface signals are multiplexed with AGP interface signals. Display Cache
interface signals only function as documented in this section when the GMCH integrated graphics is
enabled (GMCH AGP interface disabled). Refer to Section 2.11 for multiplexing map of AGP to Display
Cache interface signals.
Signal Name
Type
Description
LCS#
O
CMOS
LDQM[3:0]
O
AGP
LRAS#
O
CMOS
SDRAM Row Address Strobe. The LRAS# signal is used to generate SDRAM
Command encoded on LRAS#/LCAS#/LWE# signals. When LRAS# is sampled
active at the rising edge of the SDRAM clock, the row address is latched into the
SDRAMs.
LCAS#
O
CMOS
SDRAM Column Address Strobe. The LSCAS# signal is used to generate
SDRAM Command encoded on LSRAS#/LSCAS#/LWE# signals. When LSCAS# is
sampled active at the rising edge of the SDRAM clock, the column address is
latched into the SDRAMs.
LMA[11:0]
O
AGP
LWE#
O
CMOS
LMD[31:0]
I/O
AGP
L_FSEL
I
CMOS
Chip Select. For the memory row configured with SDRAM, this pin performs the
function of selecting the particular SDRAM components during the active state.
Input/Output Data Mask. These pins control the memory array and act as
synchronized output enables during read cycles and as a byte enables during write
cycles.
Memory Address. LMA[11:0] are used to provide the multiplexed row and column
address to SDRAM.
Write Enable Signal. LWE# is asserted during writes to SDRAM.
Memory Data. These signals are used to interface to the SDRAM data bus of
SDRAM array.
Display Cache Frequency Select. This signal indicates whether the display cache
operates at 100 MHz or 133 MHz. The value of this pin is sampled at deassertion of
CPURST# to determine display cache frequency.
HIGH = 133 MHz (Default)
LOW = 100 MHz
Note: L_FSEL has a weak internal pull-up enabled during reset.
Note: 100 MHz display cache is a non-validated feature and should be
implemented only if OEM performs validation specifically on this feature.
Datasheet
29
82815 GMCH
R
2.5.
2.6.
30
Hub Interface Signals
Signal Name
Type
Description
HL[10:0]
I/O
Hub Interface Signals. Signals used for the hub interface.
HLSTRB
I/O
Packet Strobe. One of two differential strobe signals used to transmit or receive
packet data.
HLSTRB#
I/O
Packet Strobe Compliment. One of two differential strobe signals used to transmit
or receive packet data.
HCOMP
I/O
Hub Compensation Pad. Used to calibrate the hub interface buffers. This pin
should be connected to a 40 ohm resistor tied to 1.8V VCC (VSUS_1.8)
HLREF
I
Ref
HUB Reference. Sets the differential voltage reference for the hub interface.
Display Interface Signals
Signal Name
Type
Description
VSYNC
O
3.3V
CRT Vertical Synchronization. This signal is used as the vertical sync (polarity is
programmable) or “ Vsync Interval”.
HSYNC
O
3.3V
CRT Horizontal Synchronization. This signal is used as the horizontal sync
(polarity is programmable) or “ Hsync Interval”.
IWASTE
I
Ref
Waste Reference. This signal must be tied to ground.
IREF
I
Ref
I Reference. Set pointer resistor for the internal color palette DAC.
RED
O
Analog
CRT Analog Video Output from the internal color palette DAC. The DAC is
designed for a 37.5 ohm equivalent load on each pin (e.g., 75 ohm resistor on the
board, in parallel with the 75 ohm CRT load)
GREEN
O
Analog
CRT Analog video output from the internal color palette DAC. The DAC is
designed for a 37.5 ohm equivalent load on each pin (e.g., 75 ohm resistor on the
board, in parallel with the 75 ohm CRT load)
BLUE
O
Analog
CRT Analog video output from the internal color palette DAC. The DAC is
designed for a 37.5 ohm equivalent load on each pin (e.g., 75 ohm resistor on the
board, in parallel with the 75 ohm CRT load)
DDCK
I/O
CMOS
CRT Monitor DDC Interface Clock. (Also referred to as VESATM “Display Data
Channel”, also referred to as the “Monitor Plug-n-Play” interface.) For DDC1, DDCK
and DDDA provide a unidirectional channel for Extended Display ID. For DDC2,
DDCK and DDDA can be used to establish a bi-directional channel based on I2C
protocol. The host can request Extended Display ID or Video Display Interface
information over the DDC2 channel.
DDDA
I/O
CMOS
CRT Monitor DDC Interface Data. See DDCK Description
Datasheet
82815 GMCH
R
2.7.
Digital Video Output Signals/TV-Out Pins
Signal Name
Type
Description
TVCLKIN/INT#
I
1.8V
Low Voltage TV Clock In (TV-Out Mode). In 1.8V TV-Out usage, the TVCLKIN
pin functions as a pixel clock input to the GMCH from the TV encoder. The
TVCLKIN frequency ranges from 20 MHz to 40 MHz depending on the mode
(e.g., NTSC or PAL) and the overscan compensation values in the TV Encoder.
CLKIN has a worse case duty cycle of 60%/40% coming in to the GMCH.
Flat Panel Interrupt (LCD Mode). In Flat Panel usage, the INT# pin is asserted
to cause an interrupt (typically, to indicate a hot plug or unplug of a flat panel). In
Flat Panel usage, this pin is connected internally to a pull-up resistor.
LTVCLKOUT[1:0]
O
1.8V
LCD/TV Port Clock Out: These pins provide a differential pair reference clock
that can run up to 85 MHz.
Note: It is always recommended that these pins be used as a differential pair.
Devices running at frequencies less than 65 MHz can operate in single-ended
clock mode and use LTVCLKOUT[0] as the clock. When operating in singleended clock mode, LTVCLKOUT[1] is not used.
Datasheet
LTVBLANK#
O
1.8V
Flicker Blank or Border Period Indication. BLANK# is a programmable output
pin driven by the graphics control. When programmed as a blank period
indication, this pin indicates active pixels excluding the border. When
programmed as a border period indication, this pin indicates active pixel
including the border pixels.
LTVDATA[11:0]
O
1.8V
LCD/TV Data. These signals are used to interface to the LCD/TV-Out data bus.
LTVVSYNC
O
1.8V
Vertical Sync. VSYNC signal for the LTV interface. The active polarity of the
signal is programmable.
LTVHSYNC
O
1.8V
Horizontal Sync. HSYNC signal for the LTV interface. The active polarity of the
signal is programmable.
LTVCK
I/OD
CMOS
LCD/TV Clock. Clock pin for 2-wire interface.
LTVDA
I/OD
CMOS
LCD/TV Data. Data pin for 2-wire interface.
31
82815 GMCH
R
2.8.
2.9.
32
Power Signals
Signal Name
Type
Description
V1.8
Power
Core Power (1.85V)
VDDQ
Power
AGP I/O and Display Cache Buffer Supply Power
VSUS3.3
Power
System Memory Buffer Power (Separate 3.3V power plane for power down modes)
VCCDA
Power
Display Power Signal (Connect to an isolated 1.85V plane with VCCDACA1 and
VCCDACA2)
VCCDACA1
Power
Display Power Signal
VCCBA
Power
AGP/Hub I/F Power (1.85V)
VCCDACA2
Power
Display Power Signal
VCCDPLL
Power
System Memory PLL Power (1.85V)
VSSDA
Power
Display Ground Signal
VSSDACA
Power
Display Ground Signal
VSS
Power
Core Ground
VSSDPLL
Power
System Memory PLL Ground
VSSBA
Power
AGP/Hub I/F Ground
Clock Signals
Signal Name
Type
Description
HCLK
I
CMOS
Host Clock Input. Clock used on the host interface. Externally generated
66/100/133 MHz clock.
SCLK
I
CMOS
System Memory Clock. Clock used on the output buffers of system memory.
Externally generated 100/133 MHz clock.
LTCLK[1:0]
O
CMOS
Display Cache Transmit Clocks. LTCLK[1:0] are internally generated display
cache clocks used to clock the input buffers of the SDRAM devices.
LOCLK
O
CMOS
Output Clock. LOCLK is an internally generated clock used to drive LRCLK.
LRCLK
I
CMOS
Receive Clock. LRCLK is a display cache clock used to clock the input buffers of
the GMCH.
DCLKREF
I
CMOS
Display Interface Clock. DCLKREF is a 48 MHz clock generated by an external
clock synthesizer to the GMCH.
HLCLK
I
CMOS
Hub Interface Clock. 66 MHz hub interface clock generated by an external clock
synthesizer.
RESET#
I
Global Reset. Driven by the I/O Controller Hub when PCIRST# is active.
Datasheet
82815 GMCH
R
2.10.
GMCH Power-Up/Reset Strap Options
Pin Name
Strap Description
Configuration
Interface
Type
Buffer Type
SBA[7]
Local Memory Frequency
Select
High = 133 MHz (default)
Low = 100MHz
AGP/LM
Input
SCAS#
Host Frequency
High = 133 MHz (default)
Low = 100 or 66 MHz
System
Memory
Bi-directional
SWE#
Host Frequency
High = 100 MHz (default)
Low = 66 MHz
System
Memory
Bi-directional
SMAA[11]
IOQ Depth
High = 4 (default)
Low = 1
System
Memory
Bi-directional
SMAA[10]
ALL Z
High = Normal (default)
System
Memory
Bi-directional
System
Memory
Bi-directional
System
Memory
Bi-directional
Low = All Z
SRAS#
XOR Test mode
High = Normal (default)
Low = XOR test mode
SMAA[9]
FSB P-MOS Kicker
Enable
High = enabled
(non-CuMine) (default)
Low = disabled (CuMine)
NOTES:
1. For normal operation, all strap pins must be set high “1” (except IOQ Depth and Host Frequency straps which
should be set appropriately).
2. External reset signal used to sample the straps is RESET#.
3. All system memory reset straps have internal 50K ohm pull-ups during reset.
Datasheet
33
82815 GMCH
R
2.11.
34
Multiplexed Display Cache and AGP Signal Mapping
Local Memory Signal
Name
AGP Signal Name
Local Memory Signal
Name
AGP Signal Name
LCAS#
G_AD26
LMD21
G_AD31
LCS#
G_STOP#
LMD22
SBA6
LDQM0
G_AD0
LMD23
SBA4
LDQM1
G_AD10
LMD24
PIPE#
LDQM2
SBA2
LMD25
SBA1
LDQM3
ST1
LMD26
SBA3
L_FSEL
SBA7
LMD27
G_REQ#
LMA0
G_AD22
LMD28
ST0
LMA1
G_AD15
LMD29
ST2
LMA10
G_FRAME#
LMD3
G_AD3
LMA11/LBA
G_AD18
LMD30
RBF#
LMA2
G_AD11
LMD31
SBA0
LMA3
G_BE0#
LMD4
G_AD1
LMA4
G_AD9
LMD5
G_AD6
LMA5
G_AD13
LMD6
G_AD4
LMA6
G_PAR
LMD7
G_AD2
LMA7
G_TRDY#
LMD8
G_AD12
LMA8
G_AD16
LMD9
G_AD14
LMA9
G_AD20
LRAS#
G_C/BE3#
LMD0
G_AD8
LTCLK0
G_AD30
LMD1
G_AD7
LTCLK1
G_AD28
LMD10
G_C/BE1#
LWE#
SBA5
LMD11
G_DEVSEL#
LMD12
G_IRDY#
LMD13
G_C/BE2#
LMD14
G_AD17
LMD15
G_AD19
LMD16
G_AD21
LMD17
G_AD23
LMD18
G_AD25
LMD19
G_AD27
LMD2
G_AD5
LMD20
G_AD29
Datasheet
82815 GMCH
R
2.11.1.
Display Cache Mapping at the AGP Connector
The pin mapping assignments were made with the primary goal of optimizing the layout of the AIMM
card (Display Cache add-in card that fits in the AGP connector). This was done based on the AGP
signals as they exist on the standard AGP connector. Care was taken to avoid special types of AGP
signals such as the strobes and any open-drain signals. Some signals that exist on the AGP connector that
do not exist on the GMCH’s AGP interface could not be used for Display Cache signals.
Datasheet
Pin #
B
Display Cache
Signal
1
OVRCNT#
12V
2
5.0V
TYPEDET#
3
5.0V
Reserved
4
USB+
USB-
5
GND
GND
6
INTB#
INTA#
7
CLK
RST#
8
REQ# LGM_OK
9
VCC3.3
10
ST0 LGM_OK
LMD28
ST1 LGM_OK
11
ST2 LGM_OK
LMD29
Reserved
12
RBF# LGM_OK
LMD30
PIPE# LGM_OK
13
GND
GND
14
Reserved
WBF#
15
SBA0 LGM_OK
LMD27
A
Display Cache
Signal
GNT#
VCC3.3
LMD31
SBA1 LGM_OK
LDQM2
SBA3 LGM_OK
LDQM3
LMD24
LMD25
16
VCC3.3
17
SBA2 LGM_OK
VCC3.3
18
SB_STB
SB_STB#
19
GND
GND
20
SBA4 LGM_OK
LMD23
SBA5 LGM_OK
LWE#
21
SBA6 LGM_OK
LMD22
SBA7 LGM_OK
L_FSEL
22
Reserved
Reserved
23
GND
GND
24
3.3Vaux
Reserved
25
VCC3.3
VCC3.3
26
AD31 LGM_OK
LMD21
AD30 LGM_OK
LTCLK0
27
AD29 LGM_OK
LMD20
AD28 LGM_OK
LTCLK1
28
VCC3.3
29
AD27 LGM_OK
LMD19
AD26 LGM_OK
LCAS#
30
AD25 LGM_OK
LMD18
AD24 LGM_OK
LCKE
31
GND
LMD26
VCC3.3
GND
35
82815 GMCH
R
36
Pin #
B
Display Cache
Signal
A
Display Cache
Signal
32
AD_STB1
—
AD_STB1#
—
33
AD23 LGM_OK
LMD17
C/BE3# LGM_OK
LRAS#
34
Vddq
—
Vddq
—
35
AD21 LGM_OK
LMD16
AD22 LGM_OK
LMA0
36
AD19 LGM_OK
LMD15
AD20 LGM_OK
LMA9
37
GND
—
GND
—
38
AD17 LGM_OK
LMD14
AD18 LGM_OK
LMA11
39
C/BE2# LGM_OK
LMD13
AD16 LGM_OK
LMA8
40
Vddq
—
Vddq
—
41
IRDY# LGM_OK
LMD12
FRAME# LGM_OK
LMA10
42
3.3Vaux
—
Reserved
—
43
GND
—
GND
—
44
Reserved
—
Reserved
—
45
VCC3.3
—
VCC3.3
—
46
DEVSEL# LGM_OK
LMD11
TRDY# LGM_OK
LMA7
47
Vddq
—
STOP# LGM_OK
LCS#
48
PERR#
—
PME#
—
49
GND
—
GND
—
50
SERR#
—
PAR LGM_OK
LMA6
51
C/BE1# LGM_OK
LMD10
AD15 LGM_OK
LMA1
52
Vddq
—
Vddq
—
53
AD14 LGM_OK
LMD9
AD13 LGM_OK
LMA5
54
AD12 LGM_OK
LMD8
AD11 LGM_OK
LMA2
55
GND
—
GND
—
56
AD10 LGM_OK
LDQM1
AD9 LGM_OK
LMA4
57
AD8 LGM_OK
LMD0
C/BE0# LGM_OK
LMA3
58
Vddq
—
Vddq
—
59
AD_STB0
—
AD_STB0#
—
60
AD7 LGM_OK
LMD1
AD6 LGM_OK
LMD5
61
GND
—
GND
—
62
AD5 LGM_OK
LMD2
AD4 LGM_OK
LMD6
63
AD3 LGM_OK
LMD3
AD2 LGM_OK
LMD7
64
Vddq
—
Vddq
—
65
AD1 LGM_OK
LMD4
AD0 LGM_OK
LDQM0
66
Vrefcg
—
Vrefgc
—
Datasheet
82815 GMCH
R
3.
Configuration Registers
This chapter describes the following register sets:
• PCI Configuration Registers. The GMCH contains PCI configuration registers for Device 0
(Host-hub interface Bridge/DRAM Controller), Device 1 (AGP Bridge), and Device 2 (GMCH
internal graphics device).
• Display Cache Interface Registers. This register set is used for configuration of the Display Cache
(DC) interface. The registers are located in memory space. The memory space addresses listed are
offsets from the base memory address programmed into the MMADR register (Device 2, PCI
configuration offset 14h).
• Display Cache Detect and Diagnostic Registers. This register set can be used for DC memory
detection and testing. These registers are accessed via either I/O space or memory space. The
memory space addresses listed are offsets from the base memory address programmed into the
MMADR register (Device 1, PCI configuration offset 14h).
Note that the GMCH also contains an extensive set of registers and instructions for controlling its
graphics operations. Intel graphics drivers provide the software interface at this architectural level. The
register/instruction interface is transparent at the Application Programmers Interface (API) level and
thus, beyond the scope of this document.
3.1.
Datasheet
Register Nomenclature and Access Attributes
Mnemonic
Description
RO
Read-Only. If a register is read-only, writes to this register have no effect.
R/W
Read/Write. A register with this attribute can be read and written
R/WC
Read/Write Clear. A register bit with this attribute can be read and written. However, a write of a 1
clears (sets to 0) the corresponding bit and a write of a 0 has no effect.
R/WO
Read/Write-Once. A register bit with this attribute can be written to only once after power up.
After the first write, the bit becomes read-only.
Reserved
Bits
Some of the GMCH registers described in this section contain reserved bits. These bits are
labeled “Reserved” or “Intel Reserved”. Software must deal correctly with fields that are reserved.
On reads, software must use appropriate masks to extract the defined bits and not rely on
reserved bits being any particular value. On writes, software must ensure that the values of
reserved bit positions are preserved. That is, the values of reserved bit positions must first be
read, merged with the new values for other bit positions and then written back. Note that software
does not need to perform read, merge, write operation for the configuration address register.
Reserved
Registers
In addition to reserved bits within a register, the GMCH contains address locations in the
configuration space of the Host-hub interface Bridge/DRAM Controller and the internal graphics
device entities that are marked either “Reserved” or Intel Reserved”. When a “Reserved” register
location is read, a random value can be returned. (“Reserved” registers can be 8-, 16-, or 32-bit in
size). Registers that are marked as “Reserved” must not be modified by system software. Writes
to “Reserved” registers may cause system failure.
Default
Value Upon
Reset
Upon a Full Reset, the GMCH sets all of its internal configuration registers to predetermined
default states. Some register values at reset are determined by external strapping options. The
default state represents the minimum functionality feature set required to successfully bring up
the system. Hence, it does not represent the optimal system configuration. It is the responsibility
of the system initialization software (usually BIOS) to properly determine the DRAM
configurations, operating parameters, and optional system features that are applicable, and to
program the GMCH registers accordingly.
37
82815 GMCH
R
3.2.
PCI Configuration Space Access
The GMCH and the I/O Controller Hub are physically connected via the hub interface. From a
configuration standpoint, the hub interface connecting the GMCH and the I/O Controller Hub is logically
PCI bus #0. All devices internal to the GMCH and I/O Controller Hub appear to be on PCI bus #0. The
system primary PCI expansion bus is physically attached to the I/O Controller Hub and, from a
configuration standpoint, appears as a hierarchical PCI bus behind a PCI-to-PCI bridge. The primary PCI
expansion bus connected to the I/O Controller Hub has a programmable PCI Bus number.
Note:
Even though the primary PCI expansion bus is referred to as PCI0 in this document it is not PCI bus #0
from a configuration standpoint.
The GMCH contains three PCI devices within a single physical component. The configuration registers
for Devices 0, 1, and 2 are mapped as devices residing on PCI bus #0.
• Device 0: Host-hub interface Bridge/DRAM Controller. Logically this appears as a PCI device
residing on PCI bus #0. Physically, Device 0 contains the PCI registers, DRAM registers, and other
GMCH specific registers.
• Device 1: AGP Bridge supporting 1X/2X/4X transactions. Logically this appears as a PCI device
residing on PCI bus #0.
• Device 2: GMCH internal graphics device. These registers contain the PCI registers for the GMCH
internal graphics device. Logically this appears as a PCI device residing on PCI bus #0.
Note:
3.2.1.
A physical PCI bus #0 does not exist. The hub interface and the internal devices in the GMCH and I/O
Controller Hub logically constitute PCI Bus #0 to configuration software.
PCI Bus Configuration Mechanism
The PCI Bus defines a slot based “configuration space” that allows each device to contain up to 8
functions with each function containing up to 256 8-bit configuration registers. The PCI specification
defines two bus cycles to access the PCI configuration space: Configuration Read and Configuration
Write. Memory and I/O spaces are supported directly by the processor. Configuration space is supported
by a mapping mechanism implemented within the GMCH. The PCI specification defines two
mechanisms to access configuration space, Mechanism #1 and Mechanism #2.
The GMCH supports only Mechanism #1
The configuration access mechanism makes use of the CONF_ADDR Register and CONF_DATA
Register. To reference a configuration register a DWord I/O write cycle is used to place a value into
CONF_ADDR that specifies the PCI bus, the device on that bus, the function within the device, and a
specific configuration register of the device function being accessed. CONF_ADDR[31] must be 1 to
enable a configuration cycle. CONF_DATA then becomes a window into the four bytes of configuration
space specified by the contents of CONF_ADDR. Any read or write to CONF_DATA results in the
GMCH translating the CONF_ADDR into the appropriate configuration cycle.
The GMCH is responsible for translating and routing the processor I/O accesses to the CONF_ADDR
and CONF_DATA registers to internal GMCH configuration registers, the internal graphic device, or the
hub interface.
38
Datasheet
82815 GMCH
R
3.2.2.
Logical PCI Bus #0 Configuration Mechanism
The GMCH decodes the Bus Number (bits 23:16) and the Device Number fields of the CONF_ADDR
register. If the Bus Number field of CONF_ADDR is 0, the configuration cycle is targeting a PCI Bus #0
device.
• Device #0: The Host-hub interface Bridge/DRAM Controller entity within the GMCH is hardwired
as Device #0 on PCI Bus #0.
• Device #1: The AGP interface entity within the GMCH is hardwired as Device #1 on PCI Bus #0.
• Device #2: The internal graphics device entity within the GMCH is hardwired as Device #1 on PCI
Bus #0.
Note:
3.2.3.
Configuration cycles to one of the GMCH internal devices are confined to the GMCH and not sent over
the hub interface. Note that accesses to devices #3 to #31 on PCI Bus #0 are forwarded over the hub
interface.
Primary PCI (PCI0) and Downstream Configuration Mechanism
If the Bus Number in the CONF_ADDR is non-zero, the GMCH generates a configuration cycle over the
hub interface. The I/O Controller Hub compares the non-zero Bus Number with the Secondary Bus
Number and Subordinate Bus Number registers of its P2P bridges to determine if the configuration cycle
is meant for Primary PCI expansion bus (PCI0), or a downstream PCI bus.
3.2.4.
Internal Graphics Device Configuration Mechanism
From the chipset configuration perspective, the internal graphics device is seen as a PCI device
(device 2) on PCI Bus #0. Configuration cycles that target device 2 on PCI Bus #0 are claimed by the
internal graphics device and are not forwarded via hub interface to the I/O Controller Hub.
3.2.5.
GMCH Register Introduction
The GMCH contains two sets of software accessible registers, accessed via the Host I/O address space:
• Control registers I/O mapped into the host I/O space that control access to PCI configuration space
(see section entitled I/O Mapped Registers)
• Internal configuration registers residing within the GMCH are partitioned into three logical device
register sets (“logical” since they reside within a single physical device). The first register set is
dedicated to Host-hub interface Bridge/DRAM Controller functionality (controls PCI bus 0 such as
DRAM configuration, other chip-set operating parameters, and optional features). The second
register block is dedicated to the AGP interface and the third block is dedicated to the internal
graphics device in the GMCH.
The GMCH supports PCI configuration space accesses using the mechanism denoted as Configuration
Mechanism #1 in the PCI specification.
The GMCH internal registers (both I/O Mapped and Configuration registers) are accessible by the host.
The registers can be accessed as Byte, Word (16-bit), or DWord (32-bit) quantities, with the exception of
CONF_ADDR which can only be accessed as a DWord. All multi-byte numeric fields use “little-endian”
ordering (i.e., lower addresses contain the least significant parts of the field).
Datasheet
39
82815 GMCH
R
3.3.
I/O Mapped Registers
The GMCH contains two registers that reside in the processor I/O address space − the Configuration
Address (CONF_ADDR) Register and the Configuration Data (CONF_DATA) Register. The
Configuration Address Register enables/disables the configuration space and determines what portion of
configuration space is visible through the Configuration Data window.
3.3.1.
CONF_ADDR
Configuration Address Register
I/O Address:
Default Value:
Access:
Size:
0CF8h Accessed as a DWord
00000000h
Read/Write
32 bits
CONF_ADDR is a 32 bit register accessed only when referenced as a DWord. A Byte or Word reference
will “pass through” the Configuration Address Register onto the PCI0 bus as an I/O cycle. The
CONF_ADDR register contains the Bus Number, Device Number, Function Number, and Register
Number for which a subsequent configuration access is intended.
31
30
24
CFGE
15
11
Device Number
Bit
31
23
24
Reserved (0)
10
Bus Number
8
Function Number
7
2
Register Number
1
0
Reserved
Descriptions
Configuration Enable (CFGE). This bit enables/disables accesses to PCI configuration space.
1 = Enabled.
0 = Disable.
30:24
Reserved. These bits are read-only and have a value of 0.
23:16
Bus Number. When the Bus Number is programmed to 00h the target of the Configuration Cycle is one
of the three devices in the GMCH or the PCI Bus (the hub interface is logically a PCI bus) that is directly
connected to the GMCH, depending on the Device Number field.
A type 0 Configuration Cycle is generated on the hub interface if the Bus Number is programmed to 00h
and the GMCH is not the target.
If the Bus Number is non-zero and matches the value programmed into the Secondary Bus Number
Register a Type 0 PCI configuration cycle will be generated on the AGP bridge.
If the Bus Number is non-zero, greater than the value in the Secondary Bus Number Register (Device 1)
and less than or equal to the value programmed into the Subordinate Bus Number Register (Device 1) a
Type 1 PCI configuration cycle will be generated on the AGP bridge.
If the Bus Number is non-zero, and is less than the value programmed into the Secondary Bus Number
or is greater than the value programmed into the Subordinate Bus Number Register a Type 1 hub
interface configuration cycle is generated.
40
Datasheet
82815 GMCH
R
Bit
Descriptions
15:11
Device Number. This field selects one agent on the PCI bus selected by the Bus Number. During a
Type 1 Configuration cycle, this field is mapped to AD[15:11]. During a Type 0 Configuration Cycle, this
field is decoded and one bit among AD[31:11] is driven to a 1.
The GMCH is always Device Number 0 for the Host bridge (GMCH) entity, Device Number 1 for the
AGP bridge entity, and Device Number 2 for the Internal Graphics Device entity.
If the Bus Number is non-zero and matches the value programmed into the Secondary Bus Number
Register, a Type 0 PCI configuration cycle is generated on the AGP bridge. The Device Number field is
decoded and the GMCH asserts one and only one GADxx signal as an IDSEL. GAD16 is asserted to
access Device 0, GAD17 for Device 1, GAD18 for Device 2 and so forth up to Device 15 which asserts
AD31. All device numbers higher than 15 cause a type 0 configuration access with no IDSEL asserted,
which results in a Master Abort reported in the GMCH’s “virtual” PCI-PCI bridge registers.
For Bus Numbers resulting in hub interface configuration cycles the GMCH propagates the Device
Number field as A[15:11]. For Bus Numbers resulting in AGP bridge Type 1 Configuration cycles the
Device Number is propagated as GAD[15:11].
3.3.2.
10:8
Function Number. This field is mapped to AD[10:8] during PCIx configuration cycles. This allows the
configuration registers of a particular function in a multi-function device to be accessed. The GMCH only
responds to configuration cycles with a function number of 000b; all other function number values
attempting access to the GMCH (Device Number = 0, 1 or 2, Bus Number = 0) will generate a master
abort.
7:2
Register Number. This field selects one register within a particular Bus, Device, and Function as
specified by the other fields in the Configuration Address Register. This field is mapped to AD[7:2]
during PCI configuration cycles.
1:0
Reserved.
CONF_DATA
Configuration Data Register
I/O Address:
Default Value:
Access:
Size:
0CFCh
00000000h
Read/Write
32 bits
CONF_DATA is a 32 bit read/write window into configuration space. The portion of configuration space
that is referenced by CONF_DATA is determined by the contents of CONF_ADDR.
Bit
31:0
Datasheet
Descriptions
Configuration Data Window (CDW). If bit 31 of CONF_ADDR is 1, any I/O reference that falls in the
CONF_DATA I/O space is mapped to configuration space using the contents of CONF_ADDR.
41
82815 GMCH
R
3.4.
Host-Hub Interface Bridge/DRAM Controller Device
Registers (Device 0)
Table 2 shows the GMCH configuration space for device #0.
Table 2. GMCH PCI Configuration Space (Device 0)
42
Address
Offset
Mnemonic
00–01h
VID
02–03h
DID
04–05h
Register Name
Default Value
Access
Vendor Identification
8086h
RO
Device Identification
1130h
RO
PCICMD
PCI Command
0006h
R/W
06–07h
PCISTS
PCI Status
0090h (AGP)
0080h (GFX)
RO, R/WC
08h
RID
02h (see note)
RO
09h

Reserved
00h

0Ah
SUBC
Sub-Class Code
00h
RO
0Bh
BCC
Base Class Code
06h
RO
0Ch

Reserved
00h

0Dh
MLT
Master Latency Timer
00h
RO
0Eh
HDR
Header Type
00h
RO
0Fh



10–13h
APBASE
00000008h (AGP)
00000000h (GFX)
R/W, RO
14–2Bh



2C–2Dh
SVID
Subsystem Vendor Identification
0000h
R/WO
2E–2Fh
SID
Subsystem Identification
0000h
R/WO
30–33h



34h
CAPPTR
00h (GFX)
A0h (AGP)
RO
35–4Fh



50h
GMCHCFG
GMCH Configuration
40h
R/W
51h
APCONT
Aperture Control
00h
R/WO/RO
52h
DRP
DRAM Row Population
00h
R/W
53h
DRAMT
DRAM Timing Register
00h
R/W
54h
DRP2
DRAM Row Population Register 2
00h
R/W
55–57h



58h
FDHC
Fixed DRAM Hole Control
00h
R/W
59–5Fh
PAM
Programmable Attributes Map
Registers
00h
R/W
Revision Identification
Reserved
Aperture Base Configuration
Reserved
Reserved
Capabilities Pointer
Reserved
Reserved
Datasheet
82815 GMCH
R
Address
Offset
Mnemonic
60–6Fh

70h
SMRAM
71h

72–73h
MISCC
74–87h

88–8Bh
CAPID
8C–91h

92–93h
BUFF_SC
94–95h
BUFF_SC2
96–9Fh

A0–A3h
ACAPID
A4–A7h
Register Name
Default Value
Access


00h
R/W


0000h
R/W,RO


F104A009h
RO


Buffer Strength Control
FFFFh
R/W
Buffer Strength Control 2
FFFFh
R/W


AGP Capability Identifier
00200002h
RO
AGPSTAT
AGP Status
1F000207h
RO
A8–Abh
AGPCMD
AGP Command
00000000h
R/W
AC–Afh



B0–B3h
AGPCTRL
AGP Control
00000000h
R/W
B4h
APSIZE
Aperture Size
00h
R/W
B5–B7h



B8–BBh
ATTBASE
00000000h
R/W
BCh
AMTT
AGP Multi-Transaction Timer
00h
R/W
BDh
LPTT
Low Priority Transaction Timer
00h
R/W
BEh
MCHCFG
0000 x000b
R/W, RO
BF–CAh



CBh
ERRCMD
00h
R/W
CC–FFh



Reserved
System Management RAM Control
Reserved
Miscellaneous Control Register
Reserved
Capability Identification
Reserved
Reserved
Reserved
Reserved
Aperture Translation Table Base
MCH Configuration
Reserced
Error Command
Reserved
Note: See Specification Update document for latest information.
Datasheet
43
82815 GMCH
R
3.4.1.
VID—Vendor Identification Register (Device 0)
Address Offset:
Default Value:
Attribute:
Size:
00–01h
8086h
Read-Only
16 bits
The VID Register contains the vendor identification number. This 16-bit register combined with the
Device Identification Register uniquely identify any PCI device. Writes to this register have no effect.
Bit
15:0
3.4.2.
Description
Vendor Identification Number. This is a 16-bit value assigned to Intel. Intel VID = 8086h.
DID—Device Identification Register (Device 0)
Address Offset:
Default Value:
Attribute:
Size:
02–03h
1130h
Read-Only
16 bits
This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device.
Writes to this register have no effect.
Bit
Description
15:0
Device Identification Number. This is a 16 bit value assigned to the GMCH Host-Hub Interface Bridge
/ DRAM Controller Device 0.
1130h = Device ID for Device 0.
44
Datasheet
82815 GMCH
R
3.4.3.
PCICMD—PCI Command Register (Device 0)
Address Offset:
Default:
Access:
Size:
04–05h
0006h
Read/Write
16 bits
This register provides basic control over the PCI0 interface (hub interface) ability to respond to PCI
cycles. The PCICMD Register enables and disables the SERR# signal, parity checking (PERR# signal),
GMCH’s response to PCI special cycles, and enables and disables PCI0 bus master accesses to main
memory.
15
10
Reserved (0)
9
8
FB2B
(Not Impl)
SERR En
7
6
5
4
3
2
1
0
Addr/Data
Stepping
(Not Impl)
Parity
Error En
(Not Impl)
VGA Pal
Sn
(Not Impl)
Mem WR
& Inval En
(Not Impl)
Special
Cycle En
(Not Impl)
Bus
Master En
(Not Impl)
Mem Acc
En
(Not Impl)
I/O Acc En
(Not Impl)
Bit
15:10
Descriptions
Reserved.
9
Fast Back-to-Back. (Not implemented). Hardwired to 0. Selects whether the GMCH can generate
fast back-to-back transactions to different PCI targets.
8
SERR Enable (SERRE). This bit is a global enable bit for Device 0 SERR messaging. The GMCH
does not have an SERR# signal. The GMCH communicates the SERR# condition by sending an SERR
message to the I/O Controller Hub.
1 = Enable. GMCH is enabled to generate SERR messages over the Hub interface for specific Device
0 error conditions
0 = Disable. SERR message is not generated by the GMCH for Device 0.
NOTE: This bit only controls SERR messaging for Device 0. Device 1 has its own SERRE bit to
control error reporting for error conditions occurring on Device 1. The two control bits are used
in a logical OR manner to enable the SERR hub interface message mechanism.
Datasheet
7
Address/Data Stepping. (Not implemented). Hardwired to 0.
6
Parity Error Enable (PERRE). (Not implemented). Hardwired to 0. PERR# is not implemented by
GMCH. Writes to this bit position have no affect.
5
VGA Palette Snoop. (Not implemented). Hardwired to 0. Writes to this bit position have no affect.
4
Memory Write and Invalidate Enable. The GMCH will never use this command and this bit is
hardwired to 0. Writes to this bit position will have no affects.
3
Special Cycle Enable. (Not implemented). Hardwired to 0. The GMCH ignores all special cycles
generated on the PCI.
2
Bus Master Enable (BME). (Not implemented). Hardwired to 1. The GMCH is always allowed to be a
Bus Master. . Writes to this bit position have no affect.
1
Memory Access Enable (MAE). (Not implemented). Hardwired to 1. The GMCH always allows
access to main memory. Writes to this bit position have no affect.
0
I/O Access Enable (IOAE). (Not implemented). Hardwired to 0. Writes to this bit position have no
affect.
45
82815 GMCH
R
3.4.4.
PCISTS—PCI Status Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
06–07h
0090h
Read-Only, Read/Write Clear
16 bits
PCISTS is a 16-bit status register that reports the occurrence of a PCI master abort and PCI target abort
on the PCI0 bus. PCISTS also indicates the DEVSEL# timing that has been set by the GMCH hardware
for target responses on the PCI0 bus. Bits [15:12] and bit 8 are read/write clear and bits [10:9] are readonly.
15
14
13
12
11
Detected
Par Error
(HW=0)
Sig Sys
Error
Recog
Mast Abort
Sta
Rec
Target
Abort Sta
Sig Target
Abort Sta
(HW=0)
7
6
FB2B
(HW=1)
5
Reserved
4
15
14
9
DEVSEL# Timing
(HW=00)
8
Data Par
Detected
(HW=0)
3
Cap List
(HW=1)
Bit
10
0
Reserved
Descriptions
Detected Parity Error (DPE). This bit is hardwired to a 0. Writes to this bit position have no affect.
Signaled System Error (SSE).
1 = GMCH Device 0 generates an SERR message over the hub interface for any enabled Device 0
error condition. Device 0 error conditions are enabled in the PCICMD register. Device 0 error flags
are read/reset from the PCISTS register.
0 = Software sets SSE to 0 by writing a 1 to this bit.
13
Received Master Abort Status (RMAS).
1 = GMCH generates a hub interface request that receives a Master Abort completion packet.
0 = Software clears this bit by writing a 1 to it.
12
Received Target Abort Status (RTAS).
1 = GMCH generates a hub interface request that receives a Target Abort completion packet.
0 = Software clears this bit by writing a 1 to it.
11
10:9
DEVSEL# Timing (DEVT). These bits are hardwired to 00. Writes to these bit positions have no affect.
Device 0 does not physically connect to PCI0. These bits are set to 00 (fast decode) so that optimum
DEVSEL timing for PCI0 is not limited by the GMCH.
8
Data Parity Detected (DPD). This bit is hardwired to a 0. Writes to this bit position have no affect.
7
Fast Back-to-Back (FB2B). This bit is hardwired to 1. Writes to these bit positions have no affect.
Device 0 does not physically connect to PCI. This bit is set to 1 (indicating fast back-to-back capability)
so that the optimum setting for PCI is not limited by the GMCH.
6:5
4
3:0
46
Signaled Target Abort Status (STAS). (Not implemented). Hardwired to a 0. Writes to this bit
position have no affect.
Reserved.
Capability List (CLIST). This bit is hardwired to 1 to indicate that the GMCH always has a capability
list. The list of capabilities is accessed via register CAPPTR at configuration address offset 34h.
Register CAPPTR contains an offset pointing to the address of the first of a linked list of capability
registers. Writes to this bit position have no affect.
Reserved.
Datasheet
82815 GMCH
R
3.4.5.
RID—Revision Identification Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
08h
02h (see Spec. Update document for latest information.)
Read-Only
8 bits
This register contains the revision number of the Device 0. These bits are read-only and writes to this
register have no effect.
Bit
Description
7:0
Revision Identification Number. This is an 8-bit value that indicates the revision identification number
for Device 0.
02h = A-2 Stepping
3.4.6.
SUBC—Sub-Class Code Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
0Ah
00h
Read-Only
8 bits
This register contains the Sub-Class Code for the GMCH Function #0. The register is read-only.
Bit
7:0
Description
Sub-Class Code (SUBC). This is an 8-bit value that indicates the category of Bridge into which GMCH
falls.
00h = Host Bridge.
3.4.7.
BCC—Base Class Code Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
0Bh
06h
Read-Only
8 bits
This register contains the Base Class Code of the GMCH Function #0. This register is read-only.
Bit
7:0
Description
Base Class Code (BASEC). This is an 8-bit value that indicates the Base Class Code for the GMCH.
06h = Bridge device.
Datasheet
47
82815 GMCH
R
3.4.8.
MLT—Master Latency Timer Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
0Dh
00h
Read-Only
8 bits
Device 0 is not a PCI master; therefore, this register is not implemented.
Bit
7:0
3.4.9.
Descriptions
Master Latency Timer Value. This read-only field always returns 0 when read and writes have no
affect.
HDR—Header Type Register (Device 0)
Address Offset:
Default:
Access:
Size:
0Eh
00h
Read-Only
8 bits
This register identifies the header layout of the configuration space. No physical register exists at this
location.
Bit
7:0
3.4.10.
Descriptions
Header Type. This read-only field always returns 0 when read and writes have no affect.
APBASE—Aperture Base Configuration Register
(Device 0: AGP Mode Only)
Address Offset:
Default Value (AGP Mode):
Default Value (GFX Mode):
Access:
Size:
10–3h
00000008h
00000000h
Read/Write, Read-Only
32 bits
The APBASE is a standard PCI Base Address register that is used to set the base of the AGP aperture.
The standard PCI Configuration mechanism defines the base address configuration register such that only
a fixed amount of space can be requested (dependent on which bits are hardwired to “0” or behave as
hardwired to “0”). To allow for flexibility (of the aperture) an additional register called APSIZE is used
as a “back-end” register to control which bits of the APBASE will behave as hardwired to “0”. This
register is programmed by the GMCH specific BIOS code that runs before any of the generic
configuration software is run.
Note:
48
Bit 1 of the APCONT register is used to prevent accesses to the aperture range before this register is
initialized by the configuration software and the appropriate translation table structure has been
established in the main memory.
Datasheet
82815 GMCH
R
31
26
Upper Prog. Base Address Bits
25
24
16
Lower
“HW”/Prog
Base
Address
15
4
Hardwired to 0s
Hardwired to 0s
3
Prefetch
able
2
0
Type
Mem
Space
Indicator
Bit
Description
31:26
Upper Programmable Base Address bits—R/W. These bits are used to locate the range size selected
via lower bits 25:4.
Default = 0000
25
Lower “Hardwired”/Programmable Base Address bit . This bit behaves as “hardwired” or as
programmable depending on the contents of the APSIZE register as defined below:
Aperture Size = 32 MB # r/w
Aperture Size = 64 MB # 0 (default)
Bit 25 is controlled by the bit 3 of the APSIZE register in the following manner:
• If bit APSIZE[3]=0 (indicating 64 MB aperture size), then APBASE[25]=0. If APSIZE[3]=1, then
APBASE[25]=r/w (read/write) allowing 32 MB aperture size if desired.
• Default for APSIZE[3]=0b forces default APBASE[25] = 0b (bit responds as “hardwired” to 0). This
provides a default to the maximum aperture size of 64 MB. The GMCH specific BIOS is responsible
for selecting smaller size (if required) before PCI configuration software runs and establishes the
system address map.
24:4
3
Prefetchable—RO. This bit is hardwired to 1 to identify the Graphics Aperture range as a prefetchable
(i.e., There are no side effects on reads, the device returns all bytes on reads regardless of the byte
enables, and the GMCH may merge processor writes into this range without causing errors).
2:1
Type—RO. These bits determine addressing type and they are hardwired to 00 to indicate that address
range defined by the upper bits of this register can be located anywhere in the 32-bit address space.
0
Datasheet
Hardwired to 0. This forces minimum aperture size selected by this register to be 32 MB.
Memory Space Indicator—RO. Hardwired to 0 to identify aperture range as a memory range.
49
82815 GMCH
R
3.4.11.
SVID—Subsystem Vendor Identification Register (Device 0)
Address Offset:
Default:
Access:
Size:
3.4.12.
2C–2Dh
0000h
Read/Write-Once
16 bits
Bit
Description
15:0
Subsystem Vendor ID—R/WO. This value is used to identify the vendor of the subsystem. The default
value is 0000h. This field should be programmed by BIOS during boot-up. Once written, this register
becomes read-only. This Register can only be cleared by a Reset.
SID—Subsystem Identification Register (Device 0)
Address Offset:
Default:
Access:
Size:
2E–2Fh
0000h
Read/Write-Once
16 bits
Bit
15:0
3.4.13.
Description
Subsystem ID—R/WO. This value is used to identify a particular subsystem. The default value is
0000h. This field should be programmed by BIOS during boot-up. Once written, this register becomes
read-only. This Register can only be cleared by a Reset.
CAPPTR—Capabilities Pointer (Device 0)
Address Offset:
Default Value:
Access:
Size:
34h
88h
Read-Only
8 bits
The CAPPTR provides the offset that is the pointer to the location where the capability identification
register is located.
Bit
7:0
Description
Pointer to Start of CAPPTR Linked List.
88h = Points to the CAPID register that provides capability information regarding the GMCH. The
capabilities are determined by which fuses are blown.
50
Datasheet
82815 GMCH
R
3.4.14.
GMCHCFG—GMCH Configuration Register (Device 0)
Address Offset:
Default:
Access:
Size:
50h
01ss0s00
Read/Write, Read-Only
8 bits
7
6
Mem Arb
Gnt Win
Enable
CPU
Latency
Timer
5
Reserved
4
3
2
Local
Memory
Frequency
Select
DRAM
Page
Closing
Policy
System
Memory
Frequency
Select
1
0
Reserved
Bit
Description
7
Memory Arbiter Grant Window Enable (MAGWE). This bit controls the Host vs Low Priority Graphics
timeslice regulation in the arbiter for the System DRAM.
At pre-arbitration (aka, stage 1)
0 = Disabled. Enforce fixed priority.
1 = Limit grant to host-to-graphics stream to 6 consecutive packets.
At main-arbitration (aka, stage 2)
0 = Disabled. Enforce fixed priority.
1 = 24 clocks limiting host, 24 clocks guaranteed to low priority graphics stream.
In fixed mode arbitration (MAGWE=0) the host stream always has higher priority over the low priority
graphics stream for accesses to system memory. In timeslice mode, the host stream and the low
priority graphics stream are both regulated by a time window to provide fairness to the graphics stream.
Fixed priority mode, where the host stream is always favored, is the recommended mode of operation;
this setting gives highest system performance without adversely affecting graphics performance under
real life applications workload.
6
CLT (CPU Latency Timer).
0 = Deferrable processor cycle will be Deferred immediately after receiving another ADS#
1 = Deferrable processor cycle will only be Deferred after in has been held in a “Snoop Stall” for 31
clocks and another ADS# has arrived (default).
5
Reserved.
4
Local Memory Frequency Select (LMFS). This bit selects the operating frequency for the Local
Memory Controller. Default is set by sampling the LM_FREQ_SEL strap (AGP SBA[7] pin) at reset. It
has a weak internal pull-up enabled during reset.
This is a reserved bit in the UMA Only and No Internal Graphics SKUs. A 0 is read back in these SKUs.
The output of the register bit in these SKUs is also forced to 0 such that a customer cannot effectively
program the part for 133 MHz local memory. In the Fully-Featured and 100 MHz FSB & SM SKUs,
either 1 or 0 can be programmed by the customer.
1 = 133 MHz, (default). This is a reflection of LM_FREQ_SEL strap being pulled up (default).
0 = 100 MHz. This is a reflection of LM_FREQ_SEL strap being pulled down.
Note.
Datasheet
The value of this bit should only be changed when the Internal Graphics device is disabled
(i.e., GMS = 00).
51
82815 GMCH
R
Bit
3
Description
DRAM Page Closing Policy (DPCP). When this bit is a 0, the GMCH will tend to leave the DRAM
pages open. In this mode the only times that the GMCH will close memory pages are:
0 = Precharge Bank during service of a “Page Miss” access.
Precharge All when changing from one Row to another if any Pages are open.
Precharge All at leadin to a Refresh operation
When this bit is a 1, the GMCH will tend to leave the DRAM pages closed. In this the GMCH will:
1 = Precharge All during the service of any “Page Miss” access.
Precharge All when changing from one Row to another if any Pages are open.
Precharge All at leadin to a Refresh operation.
2
System Memory Frequency Select (SMFS). This bit selects the operating frequency for the main
system memory.
Default is set by sampling SBS0# pin at reset.
0 = 100 MHz.
1 = 133 MHz.
The default Is determined by SBS0# reset strap.
1:0
52
Reserved.
Datasheet
82815 GMCH
R
3.4.15.
APCONT—Aperture Control (Device 0)
Address Offset:
Default Value:
Access:
Size:
51h
00h
Read/Write, Write-Once, Read-Only
8 bits
The Aperture Control Register controls selection and access to aperture space.
7
3
Reserved
Bit
7:3
2
2
1
0
AGP
Select
Lock
Aperture
Access
Global EN
AGP
Select
Description
Reserved.
GFX AGP Select Lock—WO. This GFX AGP Select (bit 0) can be made read-only by this bit. This is
a write-once bit. After it is written, this bit can not be changed without a system reset.
0 = GFX AGP Select remains writeable.
1 = GFX AGP Select is read-only.
1
Aperture Access Global Enable–R/W. This bit is used to prevent access to the aperture from any
port (processor, PCI0, or AGP/PCI1) before the aperture range is established by the configuration
software and appropriate translation table in the main DRAM has been initialized. It must be set after
system is fully configured for aperture accesses. Default is 0.
0
GFX AGP Select—R/W. This field selects the graphics device to be either AGP or Internal Graphics
(GFX).
0 = AGP Mode. AGP interface device is enabled. All registers in device 0 and device 1 are visible.
No device 2 registers are visible; reads from those addresses return 1s.
1 = GFX Mode. Internal Graphics device is enabled. All non-AGP related device 0 registers and all
device 2 registers are visible. No device 1 registers are visible; reads from those addresses
return 1s. Reads from AGP related device 0 registers return 0s. The internal graphics device
does not respond to any configuration cycles unless SMRAM[7:6] (@ 70h) are NOT 00 AND
APCONT[0] (@ 51h) is 1.
R/W, RO if GFX AGP Select Lock (bit 2 =1)
GFX AGP Select must be programmed before any other access is made to the configuration space.
The two possible modes are mutually exclusive. This bit determines whether other configuration
registers are enabled or disabled. This bit must be set as part of the initialization sequence. See
Software Start-Up Sequence section.
Datasheet
53
82815 GMCH
R
3.4.16.
DRP—DRAM Row Population Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
52h
00h
Read/Write (Read-Only if D_LCK = 1)
8 bits
The GMCH supports 6 physical rows of DRAM in 3 DIMMs. The width of a row is 64 bits. The DRAM
Row Population Register defines the population of each side of each DIMM. Note that this entire register
becomes read-only when the D_LCK bit is set to 1. For D_LCK bit description, see SMRAM register
(Device 0, address offset 70h).
If the system memory interface is configured to run at 133 MHz, the system BIOS must use the DRP
register (offset 52h) along with the DRP2 register (offset 52h) to detect whether the memory
configuration exceeds 2 double-sided DIMMs or 3 single-sided DIMMs. If so, the system BIOS must
down-shift the clock generator to 100 MHz to guarantee electrical integrity and timings.
7
4
3
0
DIMM 1 Population
DIMM 0 Population
Bit
Description
7:4
DIMM 1 Population. This field indicates the population of DIMM 1. (See table below)
3:0
DIMM 0 Population. This field indicates the population of DIMM 0. (See table below)
Table 3. Supported System Memory DIMM Configurations
Register
Code
DIMM
Capacity
# of Devices /
DIMM
# of
Sides
DRAM
Tech.
Front Side
Population
Count
54
N/A
Back Side
Population
Config
Count
Bank Column
Config
0
0
1
32 MB
2
32 MB
4
SS
64 Mb
4-
4 Mb x 16
3
48 MB
12
DS
64/16 Mb
4-
4
64 MB
8
DS
64 Mb
5
64 MB
8
SS
64 Mb
5
64 MB
4
SS
128 Mb
4-
8 Mb x 16
12
2
9
6
96 MB
12
DS
64 Mb
8-
8 Mb
x8
4-
4 Mb x 16
12
2
9/8
6
96 MB
8
DS
128/64 Mb
4-
8 Mb x 16
4-
4 Mb x 16
12
2
9/8
7
128 MB
16
DS
64 Mb
8-
8 Mb
x8
8-
8 Mb
x8
12
2
9
7
128 MB
8
DS
128 Mb
4-
8 Mb x 16
4-
8 Mb x 16
12
2
9
9
128 MB
8
SS
128 Mb
8-
16 Mb x 8
12
2
10
16
DS
16 Mb
Empty
Row
N/A
N/A
N/A
8-
2 Mb
x8
11
1
9
12
2
8
4 Mb x 16
8-
2 Mb
x8
12
2/1
8
4-
4 Mb x 16
4-
4 Mb x 16
8-
8 Mb
8-
2 Mb
Empty
x8
x8
A
128 MB
4
SS
256 Mb
4-
16 Mb x 16
B
192 MB
12
DS
128 Mb
8-
16 Mb x 8
4-
12
2
8
12
2
9
13
2
9
8 Mb x 16
12
2
10/9
B
192 MB
16
DS
128/64 Mb
8-
16 Mb x 8
8-
8 Mb
x8
12
2
10/9
C
256 MB
16
DS
128 Mb
8-
16 Mb x 8
8-
16 Mb x 8
12
2
10
D
256 MB
8
DS
256 Mb
4-
16 Mb x 16
4-
16 Mb x 16
13
2
9
E
256 MB
8
SS
256 Mb
8-
32 Mb x 8
13
2
10
F
512 MB
16
DS
256 Mb
8-
32 Mb x 8
13
2
10
8-
32 Mb x 8
Datasheet
82815 GMCH
R
3.4.17.
DRAMT—DRAM Timing Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
53h
00h
Read/Write
8 bits
This register controls the operating mode and the timing of the DRAM Controller.
7
5
SDRAM Mode Select
4
3
2
1
0
DRAM
Cycle
Time
Intel
Reserved
CAS#
Latency
SDRAM
RAS# to
CAS# Dly
SDRAM
RAS#
Precharge
Bit
7:5
Description
SDRAM Mode Select (SMS). These bits select the operational mode of the GMCH DRAM interface.
The special modes are intended for initialization at power up.
000 = DRAM in Self-Refresh Mode, Refresh Disabled (Default)
001 = Normal Operation, 100 MHz System memory – Refresh interval 15.6 uSec
133 MHz System memory – Refresh interval 11.7 uSec
010 = Normal Operation, 100 MHz System memory – Refresh interval 7.8
133 MHz System memory – Refresh interval 5.85 uSec
011 = Normal Operation, 100 MHz System memory – Refresh interval 1.28 uSec
133 MHz System memory – Refresh interval 0.96 uSec
100 = NOP Command Enable. In this mode all processor cycles to SDRAM result in a NOP
Command on the SDRAM interface.
101 = All Banks Precharge Enable. In this mode all processor cycles to SDRAM result in an All
Banks Precharge Command on the SDRAM interface.
110 = Mode Register Set Enable. In this mode all processor cycles to SDRAM result in a mode
register set command on the SDRAM interface. The Command is driven on the MA[12:0] lines.
MA[2:0] must always be driven to 010 for burst of 4 mode. MA3 must be driven to 1 for
interleave wrap type. MA4 needs to be driven to the value programmed in the CAS# Latency bit.
MA[6:5] should always be driven to 01. MA[12:7] must be driven to 00000. BIOS must calculate
and drive the correct host address for each row of memory such that the correct command is
driven on the MA[12:0] lines.
Note that MAB[7:4]# are inverted from MAA[7:4]; BIOS must account for this.
111 = CBR Enable. In this mode all processor cycles to SDRAM result in a CBR cycle on the SDRAM
interface.
4
DRAM Cycle Time (DCT). This bit controls the number of SCLKs for an access cycle.
0 = Tras = 5 SCLKs and Trc = 7 SCLKs (Default)
1 = Tras = 7 SCLKs and Trc = 9 SCLKs.
3
Intel Reserved.
2
CAS# Latency (CL). This bit controls the number of CLKs between when a read command is sampled
by the SDRAMs and when GMCH samples read data from the SDRAMs.
0 = CAS# latency is 3 SCLKs.
1 = CAS# latency is 2 SCLKs.
Datasheet
55
82815 GMCH
R
Bit
1
Description
SDRAM RAS# to CAS# Delay (SRCD). This bit controls the number of SCLKs from a Row Activate
command to a read or write command.
0 = 3 clocks are inserted between a row activate command and either a read or write command.
1 = 2 clocks are inserted between a row activate and either a read or write command.
0
SDRAM RAS# Precharge (SRP). This bit controls the number of SCLKs for RAS# precharge.
0 = 3 clocks of RAS# precharge are provided.
1 = 2 clocks of RAS# precharge are provided
3.4.18.
DRP2—DRAM Row Population Register 2 (Device 0)
Address Offset:
Default Value:
Access:
Size:
54h
00h
Read/Write (Read-Only if D_LCK = 1)
8 bits
This register extends support to 6 physical rows of DRAM in 3 DIMMs. The width of a row is 64 bits.
This second DRAM Row Population Register (DRP2) defines the population of each side of DIMM 2.
Note that this entire register becomes read-only when the D_LCK bit is set. For D_LCK bit description,
see SMRAM register (Device 0, address offset 70h).
If the system memory interface is configured to run at 133 MHz, the system BIOS must use the DRP
register (offset 52h) along with the DRP2 register (offset 52h) to detect whether the memory
configuration exceeds 2 double-sided DIMMs or 3 single-sided DIMMs. If so, the system BIOS must
down-shift the clock generator to 100 MHz to guarantee electrical integrity and timings.
7
4
3
Reserved
Bit
56
0
DIMM 2 Population
Description
7:4
Reserved.
3:0
DIMM 2 Population. This field indicates the population of DIMM 2. Refer to the Supported System
Memory DIMM Configurations table located with the DRP register definition. Note that some of the
larger capacity DIMMs may not be supported in DIMM 2 based on the capacities of DIMM 0 and
DIMM 1. The maximum supported main memory capacity is 512 MB.
Datasheet
82815 GMCH
R
3.4.19.
FDHC—Fixed DRAM Hole Control Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
58h
00h
Read/Write
8 bits
This 8-bit register controls a single fixed DRAM hole: 15 MB–16 MB.
7
6
0
Hole EN
Reserved
Bit
Description
7
Hole Enable (HEN). This field enables a memory hole in DRAM space. Host cycles matching an
enabled hole are passed on to the I/O Controller Hub through the hub interface. Hub interface and PCI
cycles matching an enabled hole are ignored by the GMCH. Note that a selected hole is not re-mapped.
0 = No Hole Enabled
1 = 15 MB–16 MB (1MB) Hole Enabled
6:0
3.4.20.
Reserved.
PAM—Programmable Attributes Map Registers (Device 0)
Address Offset:
Default Value:
Attribute:
Size:
59–5Fh
00h
Read/Write
4 bits/register
The GMCH allows programmable memory attributes on 13 Legacy memory segments of various sizes in
the 640 KB to 1 MB address range. Seven Programmable Attribute Map (PAM) Registers are used to
support these features. Cacheability of these areas is controlled via the MTRR registers in the P6
processor. Two bits are used to specify memory attributes for each memory segment. These bits apply to
both host, AGP/PCI, and hub interface initiator accesses to the PAM areas. These attributes are:
• Read Enable (RE). When RE = 1, the processor read accesses to the corresponding memory
segment are claimed by the GMCH and directed to main memory. Conversely, when RE = 0, the
host read accesses are directed to the hub interface/PCI0.
• Write Enable (WE). When WE = 1, the host write accesses to the corresponding memory segment
are claimed by the GMCH and directed to main memory. Conversely, when WE = 0, the host write
accesses are directed to the hub interface/PCI0.
The RE and WE attributes permit a memory segment to be Read-Only, Write Only, Read/Write, or
disabled. For example, if a memory segment has RE = 1 and WE = 0, the segment is Read-Only.
Each PAM Register controls two regions, typically 16 KB in size. Each of these regions has a 4-bit field.
The four bits that control each region have the same encoding and are defined in the following table.
Datasheet
57
82815 GMCH
R
Table 4. Attribute Bit Assignments
Bits [7, 3]
Reserved
Bits [6, 2]
Reserved
Bits [5, 1]
WE
Bits [4, 0]
RE
Description
X
X
0
0
Disabled. DRAM is disabled and all accesses
are directed to the hub interface. The GMCH
does not respond as a AGP/PCI or hub
interface target for any read or write access to
this area.
X
X
0
1
Read-Only. Reads are forwarded to DRAM
and writes are forwarded to the hub interface
for termination. This write protects the
corresponding memory segment. The GMCH
responds as a AGP/PCI or hub interface
target for read accesses but not for any write
accesses.
X
X
1
0
Write Only. Writes are forwarded to DRAM
and reads are forwarded to the hub interface
for termination. The GMCH responds as a
AGP/PCI or hub interface target for write
accesses but not for any read accesses.
X
X
1
1
Read/Write. This is the normal operating
mode of main memory. Both read and write
cycles from the host are claimed by the
GMCH and forwarded to DRAM. The GMCH
responds as a AGP/PCI or hub interface
target for both read and write accesses.
As an example, consider a BIOS that is implemented on the expansion bus. During the initialization
process, BIOS can be shadowed in main memory to increase the system performance. When BIOS is
shadowed in main memory, it should be copied to the same address location. To shadow the BIOS, the
attributes for that address range should be set to write only. BIOS is shadowed by first doing a read of
that address. This read is forwarded to the expansion bus. The host then does a write of the same address,
which is directed to main memory. After BIOS is shadowed, the attributes for that memory area are set to
read-only so that all writes are forwarded to the expansion bus. The table above and the figure below
show the PAM registers and the associated attribute bits.
58
Datasheet
82815 GMCH
R
Figure 3. PAM Registers
Offset
PAM6
PAM5
PAM4
PAM3
PAM2
PAM1
PAM0
5Fh
5Eh
5Dh
5Ch
5Bh
5Ah
59h
7
6
R
R
5
4
WE RE
3
2
R
R
1
0
WE RE
Read Enable (R/W
1=Enable
0=Disable
Reserved
Reserved
Write Enable (R/W)
1=Enable
0=Disable
Read Enable (R/W)
1=Enable
0=Disable
Write Enable (R/W)
1=Enable
0=Disable
Reserved
Reserved
pam
Table 5. PAM Registers and Associated Memory Segments
PAM Reg
Attribute Bits
PAM0[3:0]
Datasheet
Memory Segment
Comments
Reserved
Offset
59h
PAM0[7:4]
R
R
WE
RE
0F0000h–0FFFFFh
BIOS Area
59h
PAM1[3:0]
R
R
WE
RE
0C0000h–0C3FFFh
ISA Add-on BIOS
5Ah
PAM1[7:4]
R
R
WE
RE
0C4000h–0C7FFFh
ISA Add-on BIOS
5Ah
PAM2[3:0]
R
R
WE
RE
0C8000h–0CBFFFh
ISA Add-on BIOS
5Bh
PAM2[7:4]
R
R
WE
RE
0CC000h– 0CFFFFh
ISA Add-on BIOS
5Bh
PAM3[3:0]
R
R
WE
RE
0D0000h– 0D3FFFh
ISA Add-on BIOS
5Ch
PAM3[7:4]
R
R
WE
RE
0D4000h– 0D7FFFh
ISA Add-on BIOS
5Ch
PAM4[3:0]
R
R
WE
RE
0D8000h– 0DBFFFh
ISA Add-on BIOS
5Dh
PAM4[7:4]
R
R
WE
RE
0DC000h– 0DFFFFh
ISA Add-on BIOS
5Dh
PAM5[3:0]
R
R
WE
RE
0E0000h– 0E3FFFh
BIOS Extension
5Eh
PAM5[7:4]
R
R
WE
RE
0E4000h– 0E7FFFh
BIOS Extension
5Eh
PAM6[3:0]
R
R
WE
RE
0E8000h– 0EBFFFh
BIOS Extension
5Fh
PAM6[7:4]
R
R
WE
RE
0EC000h– 0EFFFFh
BIOS Extension
5Fh
59
82815 GMCH
R
DOS Area (00000h–9FFFFh)
The DOS area is 640 KB in size is always mapped to the main memory controlled by the GMCH.
Video Buffer Area (A0000h–BFFFFh)
The 128 KB graphics adapter memory region is normally mapped to a legacy video device on the hub
interface/PCI (typically VGA controller). This area is not controlled by attribute bits and processor –
initiated cycles in this region are forwarded to either the hub interface or the AGP/Internal Graphics
Device for termination. This region is also the default region for SMM space.
Accesses to this range are directed to either the hub interface or the AGP/internal Graphics Device based
on the configuration. The configuration is specified by:
1. AGP on/off configuration bit
2. AGP off: GMS bits of the SMRAM register in the GMCH Device 0 configuration space. There is
additional steering information coming from the Device 2* configuration registers and from some
of the VGA registers in the Graphics device.
3. AGP on: GMCHCFG (Device 0, bit 5, PCI-PCI Command) and BCTRL (Device 1, bit 3, PCI-PCI
Bridge Control) configuration registers
Control is applied for accesses initiated from any of the system interfaces; that is, processor bus, the hub
interface, or AGP (if enabled). Note that for the hub interface to AGP/PCI accesses, only memory write
operations are supported. Any AGP/PCI initiated VGA accesses targeting the GMCH will master abort.
For more details, see the descriptions in the configuration registers specified above.
The SMRAM Control register controls how SMM accesses to this space are treated.
Monochrome Adapter (MDA) Range (B0000h–B7FFFh)
Legacy support requires the ability to have a second graphics controller (monochrome) in the system.
In an AGP system, accesses in the standard VGA range are forwarded to the AGP bus (depending on
configuration bits). Since the monochrome adapter may be on the hub interface/PCI (or ISA) bus, the
GMCH must decode cycles in the MDA range and forward them to the hub interface. This capability is
controlled by a configuration bit (MDA bit – Device 0, BEh). In addition to the memory range B0000h
to B7FFFh, the GMCH decodes I/O cycles at 3B4h, 3B5h, 3B8h, 3B9h, 3Bah and 3BFh and forwards
them to the hub interface.
In an internal graphics system, the GMS bits of the SMRAM register in Device 0, bits in the Device 2
PCICMD register, and bits from some of the VGA registers control this functionality.
60
Datasheet
82815 GMCH
R
Expansion Area (C0000h–DFFFFh)
This 128 KB ISA Expansion region is divided into eight 16 KB segments. Each segment can be assigned
one of four Read/Write states: read-only, write-only, read/write, or disabled. Typically, these blocks are
mapped through GMCH and are subtractively decoded to ISA space. Memory that is disabled is not
remapped.
Extended System BIOS Area (E0000h–EFFFFh)
This 64 KB area is divided into four 16 KB segments. Each segment can be assigned independent read
and write attributes so it can be mapped either to main DRAM or to the hub interface. Typically, this area
is used for RAM or ROM. Memory segments that are disabled are not remapped elsewhere.
System BIOS Area (F0000h–FFFFFh)
This area is a single 64 KB segment. This segment can be assigned read and write attributes. It is by
default (after reset) read/write disabled and cycles are forwarded to the hub interface. By manipulating
the read/write attributes, the GMCH can “shadow” BIOS into the main DRAM. When disabled, this
segment is not remapped.
Datasheet
61
82815 GMCH
R
3.4.21.
SMRAM—System Management RAM Control Register
(Device 0)
Address Offset:
Default Value:
Access:
Size:
70h
00h
Read/Write, Read-Only
8 bits
The SMRAM register controls how accesses to Compatible and Extended SMRAM spaces are treated,
and how much (if any) memory is “Stolen” from the system to support both SMRAM and graphics local
memory needs.
7
6
Graphics Mode Select
Bit
7:6
5
4
Upper SMM Select
3
2
Lower SMM Select
1
0
SMM
Space
Locked
E_SMRA
M_ERR
Description
Graphics Mode Select (GMS). This field is used to enable/disable the Internal Graphics device and
select the amount of main memory that is “Stolen” to support the internal graphics device in VGA (nonlinear) mode only. These 2 bits only have meaning if we are not in AGP mode.
00 = Internal graphics device Disabled, No memory “Stolen”
01 = Internal graphics device Enabled, No memory “Stolen”
10 = Internal graphics device Enabled, 512 KB of memory “Stolen” for frame buffer.
11 = Internal graphics device Enabled, 1 MB of memory “Stolen” for frame buffer.
Notes:
• When the internal graphics device is disabled (00), the graphics device and all of its memory and I/O
functions are disabled and the clocks to this logic are turned off; memory accesses to the VGA range
(A0000–BFFFF) are forwarded on to the hub interface and the graphics local memory space is NOT
“stolen” from main memory. Any change to the SMRAM register will not affect AGP mode or cause
the controller to go into AGP mode. When this field is non-zero, the internal graphics device and all of
its memory and I/O functions are enabled; all non-SMM memory accesses to the VGA range will be
handled internally and the selected amount of graphics local memory space (0, 512 KB or 1 MB) is
“stolen” from the main memory. Graphics memory is “stolen” AFTER TSEG memory is “stolen”.
• Once D_LCK is set, these bits becomes read-only.
• GMCH does not support VGA on local memory. Software must not use the 01 mode for VGA.
62
Datasheet
82815 GMCH
R
Bit
5:4
Description
Upper SMM Select (USMM). This field is used to enable/disable the various SMM memory ranges
above 1 MB. TSEG is a block of memory (“Stolen” from Main Memory at [TOM-Size] : [TOM]) that is
only accessable by the processor and only while operating in SMM mode. HSEG is a remap of the AB
segment at FEEA0000 : FEEBFFFF. Both of these areas, when enabled, are usable as SMM RAM.
00 = TSEG and HSEG are both disabled
01 = TSEG is disabled, HSEG is conditionally enabled
10 = TSEG is enabled as 512 KB and HSEG is conditionally enabled
11 = TSEG is enabled as 1 MB and HSEG is conditionally enabled
Note:
• Non-SMM Operations (SMM processor accesses and all other access) that use these address ranges
are forwarded to the hub interface.
• Once D_LCK is set, these bits becomes read-only.
• HSEG is ONLY enabled if LSMM = 00.
3:2
Lower SMM Select (LSMM). This field controls the definition of the AB segment SMM space.
00 = AB segment disabled (no one can write to it).
01 = AB segment enabled as general system RAM (anyone can write to it).
10 = AB segment enabled as SMM Code RAM shadow. Only SMM code reads can access DRAM in the
AB segment (processor code reads only). SMM Data operations and all Non-SMM Operations go
to either the internal graphics device or are broadcast on the hub interface.
11 = AB segment enabled as SMM RAM. All SMM operations to the AB segment are serviced by
DRAM, all Non-SMM operations go to either the internal graphics device or are broadcast on the
hub interface (processor SMM R/W can access SMM space).
When D_LCK is set, bit 3 becomes read-only, and bit 2 is writable ONLY if bit 3 is a 1. When bit 3 is set,
only the processor can access it.
1
SMM Space Locked (D_LCK). When D_LCK is set to 1 then D_LCK, GMS, USMM, and the most
significant bit of LSMM become read-only. D_LCK can be set to 1 via a normal configuration space write
but can only be cleared by a reset. The combination of D_LCK and LSMM provide convenience with
security. The BIOS can use LSMM=01 to initialize SMM space and then use D_LCK to “lock down”
SMM space in the future so that no application software (or BIOS itself) can violate the integrity of SMM
space, even if the program has knowledge of the LSMM function. This bit also Locks the DRP and
DRP2 registers.
0
E_SMRAM_ERR (E_SMERR).
1 = This bit is set when processor accesses the defined memory ranges in Extended SMRAM (HSEG or
TSEG) while not in SMM mode. This bit is Not set for the case of an explicit write-back operation.
0 = It is software’s responsibility to clear this bit. Software must write a 1 to this bit to clear it.
Datasheet
63
82815 GMCH
R
3.4.22.
MISCC—Miscellaneous Control Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
72–73h
0000h
Read/Write, Read-Only
16 bits
This register holds all of the miscellaneous control bits for the GMCH .
15
14
12
SM GFX
133
Enable
7
Reserved
6
Read PWR Throttle Cntl
5
10
8
CPC Mask
4
Write PWR Throttle Cntl
Bit
15
11
Reserved
3
2
1
0
Throttle
Lock
Reserved
BNR
Looka
head
GFX LM
Win Size
Sel
Description
System Memory Graphics PC133 Enable—R/W. This bit allows the GMCH to operate in Graphics
Mode with Enhanced System Memory (PC133). Normally, SM frequency is locked to 100 MHz in
Internal Graphics mode, and GMCHCFG[2] (SMFS) is read-only. Setting this bit allows the SM
frequency to be changed by writing to GMCHCFG[2]. This bit has no effect in AGP mode.
0 = Normal Operation. GMCHCFG[2] hardwired to 0 when GMCH is in Graphics Mode
(i.e., APCONT[0] = 1)
1 = Allow 133 MHz System Memory when the GMCH is in Graphics Mode. Note that this just enables
PC133. To actually run graphics with 133 MHz SM, GMCHCFG[2] must be set to 1. Also, this bit
should be set by BIOS before GMCH is changed from AGP to Graphics mode via APCONT[0].
14
Reserved.
13
SM Transmit Stage Bypass—R/W.
0 = Normal Operation (Default). Bypass if SM=100 MHz; No bypass if SM=133 MHz.
1 = Always bypass, regardless of SM frequency.
System BIOS should set this bit to 1 to enable the bypass and optimize system memory latency by one
clock for 133 MHz operation (has no affect on 100 MHz operation).
12
Reserved.
11
CPC Mask Enable—R/W.
0 = Normal Operation (default).
1 = Never perform command per clock accesses to system memory. Mask command per clock.
Note: This bit must be set to 1 if using 133 MHz system memory.
10:8
64
Reserved.
Datasheet
82815 GMCH
R
Bit
7:6
Description
Read Power Throttle Control—R/W. These bits select the Power Throttle Bandwidth Limits for read
operations to system memory.
R/W, RO if Throttle Lock (bit 3 =1). These bits are locked (read-only) when bit 3 (Throttle Lock) is 1.
5:4
00 = No Limit
(800 MB/Sec) (Default)
01 = Limit at 87 ½ %
(700 MB/Sec)
10 = Limit at 75 %
(600 MB/Sec)
11 = Limit at 62 ½ %
(500 MB/Sec)
Write Power Throttle Control—R/W. These bits select the Power Throttle Bandwidth Limits for Write
operations to System Memory.
R/W, RO if Throttle Lock (bit 3 =1). These bits are locked (read-only) when bit 3 (Throttle Lock) is 1.
00 = No Limit
(800 MB/Sec) (Default)
01 = Limit at 62 ½ %
( 500 MB/Sec)
10 = Limit at 50 %
( 400 MB/Sec)
11 = Limit at 37 ½ %
( 300 MB/Sec)
Note: These bits must be set to ‘01’ if using 100 MHz system memory and ‘10’ if using 133 MHz
system memory.
3
Throttle Lock—R/W.
R/W, RO if Throttle Lock (bit 3 =1). Once set, this bit can only be cleared by a reset.
0 = Bits [7:3] remain writeable
1 = Block writes to bits [7:3]
2
Reserved—RO.
1
BNR Lookahead—R/W. This enables the HT unit to look further up the data path to optimize the BNR
(Block New Requests) signal to increase our effective IOQ (In Order Queue) depth.
0 = Normal Behavior (default)
1 = BNR Lookahead Enable
0
Graphics Translation Window Size Select—R/W. In GFX mode this would be the size of the GTT
(Graphics Translation Table). Not a valid bit in AGP mode.
0 = 64 MB (default)
1 = 32 MB.
Datasheet
65
82815 GMCH
R
3.4.23.
CAPID—Capability Identification (Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
88–8Bh
F104 A009h
Read-Only
32 bits
This register uniquely identifies chipset capabilities as defined in the table
below. Writes to this register have no effect.
31
30
29
28
133 MHz
Capability
Display
Cache
Capability
AGP
Capability
Internal
Graphics
Capability
27
24
CAPID Version
23
16
CAPID Length
15
8
Next Capability Pointer
7
0
CAP_ID
Bit
31
Description
133 MHz Capability—RO.
0 = Component is capable of up to 100 MHz front side bus and system memory.
1 = Component is capable of up to 133 MHz front side bus and system memory.
30
Display Cache Capability—RO.
0 = Only supports UMA mode (no local memory).
1 = Component is local memory (Display Cache) and UMA capable.
29
AGP Capability—RO.
0 = AGP mode not supported. Note that the AGP interface may still be active through the addition of
an AIMM card if bits 28 and 30 are both 1.
1 = AGP mode supported.
28
Internal Graphics Capability—RO.
0 = Internal graphic controller not supported.
1 = Internal graphic controller supported.
66
27:24
CAPID Version—RO. This field has the value 0001b to identify the first revision of the CAPID register
definition.
23:16
CAPID Length—RO. This field has the value 04h to indicate the structure length.
Datasheet
82815 GMCH
R
Bit
Description
15:8
Next Capability Pointer—RO. This field has two possible values based on APCONT[0] at offset 51h:
A0h when APCONT[0] = 0 (AGP Mode) meaning the next capability pointer is ACAPID.
00h when APCONT[0] = 1 (GFX Mode) meaning that this was the last capability pointer in the list.
CAP_ID—RO. This field has the value 1001b to identify the CAP_ID assigned by the PCI SIG for
vendor dependent capability pointers.
7:0
3.4.24.
BUFF_SC—System Memory Buffer Strength Control Register
(Device 0)
Address Offset:
Default Value:
Access:
Size:
92–93h
FFFFh
Read/Write
16 bits
This register programs the system memory DRAM interface signal buffer strengths, with the exception of
the CKEs. The programming of these bits should be based on DRAM density (x8 or x16), DRAM
technology (16Mb, 64Mb, 128Mb or 256 Mb), rows populated, etc.. Note that x4 & x32 DRAMs are not
supported. Registered DIMMs and DIMMS with ECC are also not supported and BIOS upon detection
of ECC via SPD, should report to the user that ECC DIMM timings are not supported by the GMCH.
In the descriptions below, the term “Row” is equivalent to one side of one DIMM. In other words, a
“single-sided” DIMM contains one populated row (always an odd numbered), and one empty row (even
numbered). A “double-sided” DIMM contains two populated rows.
All buffer strengths are based on the number of “loads” connected to each pin of a given signal group. A
“load” represents one pin of one SDRAM Device. The GMCH pin is implied and not counted in the load
equations. The number of loads on a given signal for a given configuration can be determined entirely
from the width of the SDRAM devices that populate each row in the configuration. This information is
readily available for each row via the Serial Presence Detect mechanism.
15
14
13
12
11
10
SCS[5]#
Buffer
Strength
SCS[4]#
Buffer
Strength
SCS[3]#
Buffer
Strength
SCS[2]#
Buffer
Strength
SCS[1]#
Buffer
Strength
SCS[0]#
Buffer
Strength
7
6
SMAB[7:4]# Buffer
Strength
Datasheet
5
4
SMAA[7:4] Buffer
Strength
3
9
2
MD and DQM Buffer
Strengths
8
SMAC[7:4]# Buffer
Strength
1
0
Control Buffer Strengths
67
82815 GMCH
R
Bit
15
Description
SCS[5]# Buffer Strength (Row 5).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
Each Row is actually selected by a pair of chip select signals (SCSA[n]# and SCSB[n]#). The number of
SCS# loads for a given row can be determined from SPD data using the following equation:
Loads = 32 / (width of SDRAM devices in row)
14
SCS[4]# Buffer Strength (Row 4).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
13
SCS[3]# Buffer Strength (Row 3).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
12
SCS[2]# Buffer Strength (Row 2).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
11
SCS[1]# Buffer Strength (Row 1).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
10
SCS[0]# Buffer Strength (Row 0).
0 = Reserved
1 = 1.0x 0 or 2 load or 4 loads
9:8
SMAC[7:4]# Buffer Strength (Rows 4/5).
00 = 2.7x > 8 loads
01 = 1.7x 8 loads
10 = 1.0x 0 or 4 loads
11 = 1.0x 0 or 4 loads
Separate copies of these SMA*[7:4] “Command-Per-Clock” signals are provided for each DIMM. So the
loads for each copy are determined by the number of SDRAM devices on the corresponding DIMM (4, 8,
12, or 16 loads). The number of loads for each SMA*[7:4] signal group can be determined from SPD
data using the following equation:
Loads = (64 / (SDRAM Device Width for 1st row)) + (64 / (SDRAM Device Width for 2nd row))
7:6
SMAB[7:4]# Buffer Strength (Rows 2/3).
00 = 2.7x > 8 loads
01 = 1.7x 8 loads
10 = 1.0x 0 or 4 loads
11 = 1.0x 0 or 4 loads
68
Datasheet
82815 GMCH
R
Bit
5:4
Description
SMAA[7:4] Buffer Strength (Rows 0/1).
00 = 2.7x > 8 loads
01 = 1.7x 8 loads
10 = 1.0x 0 or 4 loads
11 = 1.0x 0 or 4 loads
3:2
SMD[63:0] and SDQM[7:0] Buffer Strengths (All Rows).
00 = Reserved (1.7x)
01 = Reserved (0.7x)
10 = Reserved (1.0x)
11 = 1.0x 1-6 loads
The load on the SMD and SDQM signals is a function only of the number of populated rows in the
system (range 1 to 6 loads):
Loads = Number of populated rows.
1:0
SWE#, SCAS#, SRAS#, SMAA[11:8, 3:0], SBS[1:0] Control Buffer Strengths (All Rows).
00 = 1.7x > 16 loads
01 = 0.7x < 8 loads
10 = 1.0x 8-16 loads
11 = 1.0x 8-16 loads
The load on the address and control signals (other than SMA*[7:4] above) is simply the number of
devices populated in ALL rows (range from 4 to 48 loads!).
Loads = (64 / Row 0 Device Width) + (64 / Row 1 Device Width) + (64 / Row 2 Device Width) +
(64 / Row 3 Device Width) + (64 / Row 4 Device Width) + (64 / Row 5 Device Width)
Datasheet
69
82815 GMCH
R
3.4.25.
BUFF_SC2—System Memory Buffer Strength Control Register
2 (Device 0)
Address Offset:
Default Value:
Access:
Size:
94–95h
FFFFh
Read/Write
16 bits
This register programs the system memory DRAM interface CKE signal buffer strengths. See BUFF_SC
register for the remainder of the buffer strength controls.
15
8
Reserved (R/W)
7
6
Reserved (R/W)
5
4
3
2
1
0
CKE5
Buffer
Strength
CKE4
Buffer
Strength
CKE3
Buffer
Strength
CKE2
Buffer
Strength
CKE1
Buffer
Strength
CKE0
Buffer
Strength
Bit
15:6
5
Description
Reserved.
SCKE[5] Buffer Strength (Row 5).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
The load on a given SCKE signal is equal to the number of SDRAM devices for that particular row
(either 4 or 8 loads).
Loads = (64 / SDRAM Device Width for this row)
4
SCKE[4] Buffer Strength (Row 4).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
3
SCKE[3] Buffer Strength (Row 3).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
2
SCKE[2] Buffer Strength (Row 2).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
1
SCKE[1] Buffer Strength (Row 1).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
0
SCKE[0] Buffer Strength (Row 0).
0 = 2.7x 8 loads
1 = 1.7x 0 or 4 loads
70
Datasheet
82815 GMCH
R
3.4.26.
SM_RCOMP—System Memory R Compensation Control
Register (Device 0)
Address Offset:
Default Value:
Access:
Size:
98–9Bh
XXXXXXXXh
Read/Write, Read-Only
32 bits
This register controls the system memory Rcomp buffers (both horizontally and vertically oriented).
31
30
15
23
22
Reserved
V Override
Enable
14
H Override
Enable
7
Reserved
20
SRCOMP_VP
6
4
SRCOMP_HP
Bit
31
19
18
Reserved
3
Reserved
16
SRCOMP_VN
2
0
SRCOMP_HN
Description
SRCOMP_V Override Enable—R/W.
0 = SM Rcomp is active for vertically oriented buffers (Default).
1 = SM Rcomp is NOT-active for vertically oriented buffers.
30:23
Reserved.
22:20
SRCOMP_VP—RO or R/W. P-Channel Compensation Value for Vertical Buffers. This value is
generated by the Rcomp logic to control the drive characteristics of the vertically oriented P-channel
devices in the SM buffers.
In Normal operation, field is read-only and reflects current compensation.
In Override Mode (see bit 31), field is written with desired compensation value which is loaded via
software when SM Rcomp operation is disabled.
19
18:16
Reserved.
SRCOMP_VN—RO or R/W. N-Channel Compensation Value for Vertical Buffers. This value is
generated by the Rcomp logic to control the drive characteristics of the vertically oriented N-channel
devices in the SM buffers.
In Normal operation, field is read-only and reflects current compensation.
In Override Mode (see bit 31), field is written with desired compensation value which is loaded via
software when SM Rcomp operation is disabled.
15
SRCOMP_H Override Enable—R/W.
0 = SM Rcomp is active for horizontally oriented buffers (Default).
1 = SM Rcomp is NOT-active for horizontally oriented buffers.
14:7
Datasheet
Reserved.
71
82815 GMCH
R
Bit
6:4
Description
SRCOMP_HP—RO or R/W. P-Channel Compensation Value for Horizontal Buffers. This value is
generated by the Rcomp logic to control the drive characteristics of the horizontally oriented P-channel
devices in the SM buffers.
In Normal operation, field is read-only and reflects current compensation.
In Override Mode (see bit 15), field is written with desired compensation value which is loaded via
software when SM Rcomp operation is disabled.
3
Reserved.
2:0
SRCOMP_HN—RO or R/W. N-Channel Compensation Value for Horizontal Buffers. This value is
generated by the Rcomp logic to control the drive characteristics of the horizontally oriented N-channel
devices in the SM buffers.
In Normal operation, field is read-only and reflects current compensation.
In Override Mode (see bit 15), field is written with desired compensation value which is loaded via
software when SM Rcomp operation is disabled.
3.4.27.
SM—System Memory Control Register
Address Offset:
Default Value:
Access:
Size:
9C–9Fh
XXXXXXXXh
Read/Write, Read-Only
32 bits
This register controls the two System Memory Delay Locked Loop (DLL) blocks that offset the transmit
and receive clocks used to interface with the external SDRAM devices. The Transmit DLL provides an
early version of SCLK to provide additional setup margin to the external SDRAM devices. The Receive
DLL provides a late version of SCLK to provide additional setup time on read data driven by the
SDRAM devices back to the GMCH.
By default, the Transmit DLL is enabled (whether the operating frequency is 100 MHz or 133 MHz).
The Receive DLL is always bypassed, regardless of operating frequency. When the RDLL is bypassed,
the RDLL Bias field, instead, controls a buffer delay chain with programmable tap points. This chain has
8 tap points each with approximately 200 ps of incremental delay at the slow corner (total delay range
0 to 1.4 ns) and about 80 ps of incremental delay at the “fast” corner (0 to 0.56 ns total range).
31
16
Reserved
Bit
15
TDLL
Bypass
14
0
Reserved
Description
31:16
Reserved.
15
Transmit DLL Enable (TDLLE)—R/W.
0 = TDLL Enabled (Default)
1 = TDLL Disabled and bypassed
14:0
72
Reserved.
Datasheet
82815 GMCH
R
3.4.28.
ACAPID—AGP Capability Identifier Register
(Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
A0–A3h
00200002h
Read-Only
32 bits
This register provides standard identifier for AGP capability.
31
24
Reserved
23
20
19
Major AGP Revision Number
16
Minor AGP Revision Number
15
8
Next Capability Pointer
7
0
AGP Capability ID
Bit
Description
31:24
Reserved.
23:20
Major AGP Revision Number. These bits provide a major revision number of AGP specification that
this version of GMCH conforms. These bits are set to the value 0010b to indicate AGP Rev. 2.x.
19:16
Minor AGP Revision Number. These bits provide a minor revision number of AGP specification that
this version of GMCH conforms. This number is hardwired to a value of “0000” (i.e., implying Rev x.0).
Together with major revision number this field identifies GMCH as an AGP REV 2.0 compliant device.
Datasheet
15:8
Next Capability Pointer. AGP capability is the first and the last capability described via the capability
pointer mechanism; therefore, these bits are hardwired to 0s to indicate the end of the capability linked
list.
7:0
AGP Capability ID. This field identifies the linked list item as containing AGP registers. This field has
the value 0000_0010b as assigned by the PCI SIG.
73
82815 GMCH
R
3.4.29.
AGPSTAT—AGP Status Register (Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
A4–A7h
1F000207h
Read-Only
32 bits
This register reports AGP device capability/status.
31
24
Request Queue (RQ) (HW=1Fh)
23
16
Reserved
15
10
Reserved
7
6
Reserved
Bit
31:24
5
4
3
>4 GB
Support
(HW=0)
Fast
Writes
(HW=0)
Reserved
1
8
SBA
(HW=1)
Reserved
2
0
Data Transfer Rate
(HW=111; 1x,2x,4x modes supported)
Description
Request Queue (RQ). This field is hardwired to 1Fh to indicate a maximum of 32 outstanding AGP
command requests can be handled by the GMCH. This field contains the maximum number of AGP
command requests the GMCH is configured to manage. The lower 6 bits of this field reflect the value
programmed in AGPCTRL[12:10]. Only discrete values of 32, 16, 8, 4 , 2 and 1 can be selected via
AGPCTRL. Upper bits are hardwired to 0.
Default =1Fh to allow a maximum of 32 outstanding AGP command requests.
23:10
9
8:6
Reserved
SideBand Addressing (SBA). Indicates the GMCH supports sideband addressing. Hardwired to 1.
Reserved
5
Greater Than 4 GB Address Support (4GB). This bit indicates that the GMCH does not support
addresses greater than 4 GB. It is hardwired to 0.
4
Fast Writes (FW). This bit indicates that the GMCH does not support Fast Writes from the processor
to the AGP master. It is hardwired to a 0.
3
Reserved
2:0
Data Transfer Rate Capability (RATE). After reset the GMCH reports its data transfer rate capability.
Note that the selected data transfer mode applies to both AD bus and SBA bus.
Bit 0 = 1 = 1x data transfer mode
Bit 1 = 1 = 2x data transfer mode
Bit 2 = 1 = 4x data transfer mode. This bit can be masked by the AGPCTRL register bit 0
(AGP 4X Override).
1x , 2x , and 4x data transfer modes are supported by the GMCH; therefore, this bit field has a Default
Value = 111.
74
Datasheet
82815 GMCH
R
3.4.30.
AGPCMD—AGP Command Register (Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
A8–Abh
00000000h
Read/Write
32 bits
This register provides control of the AGP operational parameters.
31
10
Reserved
7
6
Reserved
5
4
3
4 GB
(HW=0)
FW EN
Reserved
Bit
31:10
9
9
8
SBA EN
AGP EN
2
0
Data Rate
Description
Reserved.
Sideband Address Enable (SBA).
1 = Enable. The sideband addressing mechanism is enabled.
0 = Disable
8
7:6
AGP Enable. When this bit is reset to 0, the GMCH ignores all AGP operations, including the sync
cycle. Any AGP operations received while this bit is set to 1 will be serviced, even if this bit is reset to
0. If this bit transitions from a 1 to a 0 on a clock edge in the middle of an SBA command being
delivered in 1X mode, the command is issued. When this bit is set to 1 the GMCH will respond to AGP
operations delivered via PIPE#, or to operations delivered via SBA, if the AGP Side Band Enable bit is
also set to 1.
Reserved.
5
Greater Than 4 GB Support (4GB). Hardwired to 0. The GMCH as an AGP target does not support
addressing greater than 4 GB.
4
Fast Writes Enable (FW). This bit must always be programmed to 0. The chipset will behave
unpredictably if this bit is programmed with 1.
3
Reserved.
2:0
Data Rate Capability. The settings of these bits determines the AGP data transfer rate. One (and only
one) bit in this field must be set to indicate the desired data transfer rate. The same bit must be set on
both master and target. Configuration software will update this field by setting only one bit that
corresponds to the capability of AGP master (after that capability has been verified by accessing the
same functional register within the AGP master’s configuration space.)
Bit 0 = 1= 1X
Bit 1 = 1 = 2X
Bit 2 = 1 = 4x
Bit 2 becomes reserved (but will still read 4x, erroneously) when the 4x Override bit in the AGP CTRL
register is set to 1 because this bit will not be updated in 4x Override mode. When the 4x Override bit
is set writes to Data Rate[2] have no functional impact.
Note: This field applies to AD and SBA buses.
Datasheet
75
82815 GMCH
R
3.4.31.
AGPCTRL—AGP Control Register (Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
B0–B3h
00000000h
Read/Write
32 bits
This register provides for additional control of the AGP interface.
31
8
Reserved
Bit
31:8
7
7
GTLB_EN
6
1
Reserved
0
4X
Override
Description
Reserved
GTLB Enable ( and GTLB Flush Control)—R/W.
1 = Enables normal operations of the Graphics Translation Lookaside Buffer.
0 = Disable (default). The GTLB is flushed by clearing the valid bits associated with each entry. In this
mode of operation all accesses that require translation bypass the GTLB. All requests that are
positively decoded to the graphics aperture force the GMCH to access the translation table in main
memory before completing the request. Translation table entry fetches are not cached in the GTLB.
NOTE:
• When an invalid translation table entry is read, this entry is still cached in the GTLB (ejecting the least
recently used entry).
• The GMCH flushes the GWB when software sets or clears this bit to ensure coherency between the
GTLB and main memory.
• This bit can be changed dynamically (i.e., while an access to GTLB occurs).
6:1
0
Reserved
4X Override. When this bit is set to 1 the Rate[2] bit in the AGPSTAT register will be read as a 0. This
“back-door” register bit allows BIOS to disable AGP 4X mode.
The introduction of universal AGP cards and universal motherboards has raised some potential
problems that this bit alleviates. AGP 2X can operation at 1.5V or 3.3V. AGP 4X can operate only at
1.5V. In a system that is supporting 3.3V operation, and therefore cannot support a 4X transfer
rate, it is the responsibility of the BIOS to make sure that 4X mode is not selected.
76
Datasheet
82815 GMCH
R
3.4.32.
APSIZE—Aperture Size (Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
B4h
00h
Read/Write
8 bits
This register determines the effective size of the graphics aperture used for a particular GMCH
configuration. This register can be updated by the GMCH-specific BIOS configuration sequence before
the PCI standard bus enumeration sequence takes place. If the register is not updated, a default value will
select an aperture of maximum size (i.e., 64 MB).
7
4
Reserved
Bit
7:4
3
3
GFX
Aperture
Size
2
0
Reserved
Description
Reserved.
Graphics Aperture Size (GASIZE). Bit 3 operates on bit 25 of the Aperture Base (APBASE)
configuration register. When this bit is 0, it forces bit 25 in APBASE to behave as “hardwired” to 0.
When this bit is 1, it forces bit 25 in APBASE to be read/write accessible. Only the following
combinations are allowed:
0 = 64 MB Aperture Size
1 = 32 MB Aperture Size
Default for APSIZE[3]=0b forces default APBASE[25] =0b (responds as “hardwired” to 0). This provides
maximum aperture size of 64 MB. Programming APSIZE[3]=1b enables APBASE[25] as read/write
programmable.
2:0
Datasheet
Reserved.
77
82815 GMCH
R
3.4.33.
ATTBASE—Aperture Translation Table Base Register
(Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
B8–BBh
00000000h
Read/Write
32 bits
This register provides the starting address of the Graphics Aperture Translation Table Base located in the
main DRAM. This value is used by the GMCH’s Graphics Aperture address translation logic (including
the GTLB logic) to obtain the appropriate address translation entry required during the translation of the
aperture address into a corresponding physical DRAM address. The ATTBASE register may be
dynamically changed.
Note:
The address provided via ATTBASE is 4 KB aligned.
31
29
Reserved
Bit
78
28
12
ATT Base Address
11
0
Reserved
Description
31:29
Reserved.
28:12
ATT Base Address. This field contains a pointer to the base of the translation table used to map
memory space addresses in the aperture range to addresses in main memory.
11:0
Reserved.
Datasheet
82815 GMCH
R
3.4.34.
AMTT—AGP Multi-Transaction Timer
(Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
BCh
00h
Read/Write
8 bits
AMTT controls the amount of time that the GMCH’s arbiter allows AGP/PCI master to perform multiple
back-to-back transactions. The GMCH’s AMTT mechanism is used to optimize the performance of the
AGP master (using PCI semantics) that performs multiple back-to-back transactions to fragmented
memory ranges (and as a consequence it can not use long burst transfers). The AMTT mechanism applies
to the processor — AGP/PCI transactions as well and it guarantees to the processor a fair share of the
AGP/PCI interface bandwidth.
The number of clocks programmed in the AMTT represents the guaranteed time slice (measured in
66 MHz clocks) allotted to the current agent (either AGP/PCI master or Host bridge) after which the
AGP arbiter grants the bus to another agent. The default value of AMTT is 00h and disables this
function. The AMTT value can be programmed with 8-clock granularity. For example, if the AMTT is
programmed to 18h, the selected value corresponds to the time period of 24 AGP (66 MHz) clocks.
7
3
Multi-Transaction Timer Count Value
Bit
Datasheet
2
0
Reserved
Description
7:3
Multi-Transaction Timer Count Value. The number programmed in these bits represents the
guaranteed time slice (measured in eight 66 MHz clock granularity) allotted to the current agent (either
AGP/PCI master or Host bridge) after which the AGP arbiter will grant the bus to another agent.
2:0
Reserved.
79
82815 GMCH
R
3.4.35.
LPTT—AGP Low Priority Transaction Timer Register
(Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
BDh
00h
Read/Write
8 bits
LPTT is similar in function to AMTT. This register is used to control the minimum tenure on the AGP
for low priority data transaction (both reads and writes) issued using PIPE# or SB mechanisms.
The number of clocks programmed in the LPTT represents the guaranteed time slice (measured in
66 MHz clocks) allotted to the current low priority AGP transaction data transfer state. This does not
necessarily apply to a single transaction but it can span over multiple low-priority transactions of the
same type. After this time expires, the AGP arbiter may grant the bus to another agent if there is a
pending request. The LPTT does not apply in the case of high-priority request where ownership is
transferred directly to high-priority requesting queue. The default value of LPTT is 00h and disables this
function. The LPTT value can be programmed with 8-clock granularity. For example, if the LPTT is
programmed to 10h, the selected value corresponds to the time period of 16 AGP (66 MHz) clocks.
7
3
Low Priority Transaction Timer Count Value
80
2
0
Reserved
Bit
Description
7:3
Low Priority Transaction Timer Count Value. The number of clocks programmed in these bits
represents the guaranteed time slice (measured in eight 66 MHz clock granularity) allotted to the current
low priority AGP transaction data transfer state.
2:0
Reserved.
Datasheet
82815 GMCH
R
3.4.36.
GMCHCFG—GMCH Configuration Register
(Device 0: AGP Mode Only)
Address Offset:
Default:
Access:
Size:
BEh
0000 X000b
Read/Write, Read-Only
8 bits
7
6
Reserved
5
4
3
MDA
Present
(R/W)
Reserved
AGP_BUF
Mode
(RO)
Bit
7:6
5
2
0
Reserved
Description
Reserved.
MDA Present (MDAP)—R/W. This bit works with the VGA Enable bit in the BCTRL register (3Eh, bit 3)
of device 1 to control the routing of processor-initiated transactions targeting MDA compatible I/O and
memory address ranges. This bit should not be set when the VGA Enable bit is not set. If the VGA
enable bit is set, accesses to IO address range x3BCh–x3BFh are forwarded to the hub interface. If the
VGA enable bit is not set, accesses to IO address range x3BCh–x3BFh are treated just like any other IO
accesses (i.e., the cycles are forwarded to AGP if the address is within IOBASE and IOLIMIT, and the
ISA enable bit is not set; otherwise, they are forwarded to the hub interface). MDA resources are defined
as the following:
• Memory:
0B0000h–0B7FFFh
• I/O:
3B4h, 3B5h, 3B8h, 3B9h, 3BAh, 3BFh,
(including ISA address aliases, A[15:10] are not used in decode)
Any I/O reference that includes the I/O locations listed above, or their aliases, will be forwarded to the
hub interface, even if the reference includes I/O locations not listed above.
The following table shows the behavior for all combinations of MDA and VGA:
VGA
MDA
Behavior
0
0
All references to MDA and VGA go to hub interface
0
1
Illegal combination (DO NOT USE)
1
0
All references to VGA go to AGP/PCI. MDA-only references (I/O address 3BFh and
aliases) will go to the hub interface.
1
1
VGA references go to AGP/PCI; MDA references go to the hub interface.
4
Reserved.
3
AGP I/O Buffer Mode (AGP_BUF)—RO. The GMCH has an internal circuit that detects the voltage
level on the AGP I/O buffer VDDQ rail. The voltage level information is latched 500 us after the
deasserting edge of RSTIN# and stored in this register bit.
1 = AGP VDDQ is sensed at 3.3V.
0 = AGP VDDQ is sensed at 1.5V.
2:0
Datasheet
Reserved.
81
82815 GMCH
R
3.4.37.
ERRCMD—Error Command Register
(Device 0: AGP Mode Only)
Address Offset:
Default Value:
Access:
Size:
CBh
00h
Read/Write
8 bits
This register enables various errors to generate a SERR hub interface special cycle. Since the GMCH
does not have an SERR# signal, SERR messages are passed from the GMCH to the I/O Controller Hub
over the hub interface. The actual generation of the SERR message is globally enabled for Device 0 via
the PCI Command register.
Note:
An error can generate one and only one hub interface error special cycle. It is software’s responsibility to
make sure that when an SERR error message is enabled for an error condition, SMI and SCI error
messages are disabled for that same error condition.
7
6
Reserved
5
4
3
2
1
0
SRMMRO
SRTA
SLNDM
SAAOGA
SIAA
SAIGATTE
Bit
7:6
5
Description
Reserved.
SERR on Receiving Main Memory Refresh Overrun Enable. Identical functionality in Device 2
memory mapped space @ 020B8h. This bit allows use of this same functionality in AGP Mode.
1 = Enable. GMCH generates a SERR hub interface special cycle when a main memory refresh
overrun occurs.
0 = Disable. Reporting of this condition is disabled.
4
SERR on Receiving Target Abort on the hub interface.
1 = Enable. GMCH generates a SERR hub interface special cycle when a GMCH-originated hub
interface cycle is terminated with a Target Abort.
0 = Disable. Reporting of this condition is disabled.
3
SERR on LOCK to non-DRAM Memory.
1 = Enable. GMCH generates a SERR hub interface special cycle when a processor-initiated LOCK
transaction targeting non-DRAM memory space occurs.
0 = Disable. Reporting of this condition is disabled.
2
SERR on AGP Access Outside of Graphics Aperture.
1 = Enable. GMCH generates a SERR hub interface special cycle when an AGP access occurs to an
address outside of the graphics aperture.
0 = Disable. Reporting of this condition is disabled.
1
SERR on Invalid AGP Access.
1 = Enable. GMCH generates a SERR hub interface special cycle when an AGP access occurs to an
address outside of the graphics aperture and either to the 640 KB–1 MB range or above the top of
memory.
0 = Disable.
0
SERR on Access to Invalid Graphics Aperture Translation Table Entry.
1 = Enable. GMCH generates a SERR hub interface special cycle when an invalid translation table
entry was returned in response to a AGP access to the graphics aperture.
0 = Disable. Reporting of this condition via SERR messaging is disabled.
82
Datasheet
82815 GMCH
R
Table 6. Summary of GMCH Error Sources, Enables and Status Flags
Error Event
Hub Interface
Message
Processor LOCK to non-DRAM
memory
SERR
Received Hub Interface Target Abort
SERR
Enable Bits Required
to be Set
PCICMD bit 8
Status Flags Set
PCISTS bit 14
ERRCMD bit 3
AGP Access Outside of Graphics
Aperture
SERR
Invalid AGP Access
SERR
PCICMD bit 8
PCISTS bit 14
ERRCMD bit 4
PCISTS bit 12
PCICMD bit 8
PCISTS bit 14
ERRCMD bit 2
PCICMD bit 8
PCISTS bit 14
ERRCMD bit 1
Access to Invalid GTLB Entry
SERR
PCICMD bit 8
PCISTS bit 14
ERRCMD bit 0
AGP/PCI Parity Error Detected
AGP/PCI Received Target Abort
Datasheet
SERR
SERR
PCICMD1 bit 8
PCISTS1 bit 14
BCTRL bit 2
PCISTS1 bit 15
PCICMD1 bit 8
PCISTS1 bit 14
ERRCMD1 bit 0
PCISTS1 bit 12
83
82815 GMCH
R
3.5.
AGP/PCI Bridge Registers
(Device 1: Visible in AGP Mode Only)
These registers are accessible through the configuration mechanism defined in an earlier section of this
document.
Table 7. GMCH Configuration Space (Device 1)
Address
Offset
Mnemonic
Register Name
Default Value
Access
Type
00–01h
VID1
Vendor Identification
8086h
RO
02–03h
DID1
Device Identification
1131h
RO
04–05h
PCICMD1
PCI Command
0000h
RO, R/W
06–07h
PCISTS1
PCI Status
0020h
RO, R/WC
08
RID1
02h (see note)
RO
09

Reserved
00h

0Ah
SUBC1
Sub-Class Code
04h
RO
0Bh
BCC1
Base Class Code
06h
RO
0Ch

Reserved
00h

0Dh
MLT1
Master Latency Timer
00h
R/W
0Eh
HDR1
Header Type
01h
RO
0F–17h

Reserved
00h

18h
PBUSN
Primary Bus Number
00h
RO
19h
SBUSN
Secondary Bus Number
00h
R/W
1Ah
SUBUSN
Subordinate Bus Number
00h
R/W
Revision Identification
1Bh
SMLT
Secondary Bus Master Latency Timer
00h
R/W
1Ch
IOBASE
I/O Base Address
F0h
R/W
1Dh
IOLIMIT
I/O Limit Address
00h
R/W
1E–1Fh
SSTS
Secondary Status
02A0h
RO, R/WC
20–21h
MBASE
Memory Base Address
FFF0h
R/W
22–23h
MLIMIT
Memory Limit Address
0000h
R/W
24–25h
PMBASE
Prefetchable Memory Base Address
FFF0h
R/W
26–27h
PMLIMIT
Prefetchable Memory Limit Address
0000h
R/W
28–3Dh

Reserved
00h

3Eh
BCTRL
Bridge Control
00h
R/W
3Fh

Reserved
00h

40h
ERRCMD1
Error Command
00h
R/W
41–FFh

Reserved
00h

Note: See Specification Update document for latest information.
84
Datasheet
82815 GMCH
R
3.5.1.
VID1—Vendor Identification Register (Device 1)
Address Offset:
Default Value:
Attribute:
Size:
00–01h
8086h
Read-Only
16 bits
The VID1 Register contains the vendor identification number. This 16-bit register combined with the
Device Identification Register uniquely identify any PCI device. Writes to this register have no effect.
Bit
Description
Vendor Identification Number. This is a 16-bit value assigned to Intel. Intel VID = 8086h.
15:0
3.5.2.
DID1—Device Identification Register (Device 1)
Address Offset:
Default Value:
Attribute:
Size:
02–03h
1131h
Read-Only
16 bits
This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device.
Writes to this register have no effect.
Bit
Description
Device Identification Number. This is a 16 bit value assigned to the GMCH AGP interface device.
15:0
1131h = Device ID for Device 1.
3.5.3.
PCICMD1—PCI-PCI Command Register (Device 1)
Address Offset:
Default:
Access:
Size
04–05h
0000h
Read/Write, Read-Only
16 bits
15
10
Reserved (0)
Datasheet
9
8
FB2B
(Not Impl)
SERR En
7
6
5
4
3
2
1
0
Addr/Data
Stepping
(Not Impl)
Parity
Error En
(Not Impl)
Reserved
Mem WR
& Inval En
(Not Impl)
Special
Cycle En
(Not Impl)
Bus
Master En
Mem Acc
En
I/O Acc En
85
82815 GMCH
R
Bit
15:10
Descriptions
Reserved.
9
Fast Back-to-Back. (Not Applicable). Hardwired to 0.
8
SERR Message Enable (SERRE1). This bit is a global enable bit for Device 1 SERR messaging. The
GMCH does not have an SERR# signal. The GMCH communicates the SERR# condition by sending
an SERR message to the I/O Controller Hub. If this bit is set to a 1, the GMCH is enabled to generate
SERR messages over the hub interface for specific Device 1 error conditions that are individually
enabled in the ERRCMD1 and BCTRL registers. The error status is reported in the PCISTS1 register. If
SERRE1 is reset to 0, the SERR message is not generated by the GMCH for Device 1.
1 = Enable.
0 = Disable.
NOTE: This bit only controls SERR messaging for the Device 1. Device 0 has its own SERRE bit to
control error reporting for error conditions occurring on Device 0. The two control bits are used
in a logical OR manner to enable the SERR hub interface message mechanism.
7
Address/Data Stepping. (Not Applicable). Hardwired to 0.
6
Parity Error Enable (PERRE1). Hardwired to 0. PERR# is not supported on AGP/PCI1.
5
Reserved.
4
Memory Write and Invalidate Enable—RO. This bit is implemented as read-only and returns a value
of 0 when read.
3
Special Cycle Enable—RO. This bit is implemented as read-only and returns a value of 0 when read.
2
Bus Master Enable (BME1)—R/W.
1
1=
Enable. AGP Master-initiated FRAME# cycles are accepted by the GMCH if they hit a valid
address decode range. This bit has no affect on AGP Master originated SBA or PIPE# cycles.
0=
Disable (default). AGP Master-initiated FRAME# cycles are ignored by the GMCH resulting in a
Master Abort. Ignoring incoming cycles on the secondary side of the P2P bridge effectively
disables the bus master on the primary side.
Memory Access Enable (MAE1)—R/W.
1 = Enable. Enables the Memory and Prefetchable memory address ranges defined in the MBASE,
MLIMIT, PMBASE, and PMLIMIT registers, as well as the VGA window.
0 = Disable. All of the memory space for Device 1 is disabled.
0
I/O Access Enable (IOAE1)—R/W.
1 = Enable. Enables the I/O address range defined in the IOBASE and IOLIMIT registers.
0 = Disable. All of I/O space for Device 1 is disabled.
86
Datasheet
82815 GMCH
R
3.5.4.
PCISTS1—PCI-PCI Status Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
06–07h
0020h
Read-Only, Read/Write Clear
16 bits
PCISTS1 reports the occurrence of error conditions associated with the primary side of the “virtual”
PCI-PCI bridge embedded in the GMCH. Since this device does not physically reside on PCI0, it reports
the optimum operating conditions so that it does not restrict the capability of PCI0.
15
14
13
12
11
Detected
Par Error
(HW=0)
Sig Sys
Error
Recog
Mast Abort
Sta
(HW=0)
Rec
Target
Abort Sta
(HW=0)
Sig Target
Abort Sta
(HW=0)
7
6
FB2B
(HW=1)
5
Reserved
Bit
4
10
9
DEVSEL# Timing
(HW=00)
3
Cap List
(HW=1)
8
Data Par
Detected
(HW=0)
0
Reserved
Descriptions
15
Detected Parity Error (DPE1). (Not Applicable). Hardwired to 0.
14
Signaled System Error (SSE1).
1 = GMCH Device 1 generated an SERR message over hub interface for any enabled Device 1 error
condition. Device 1 error conditions are enabled in the PCICMD1, ERRCMD1 and BCTRL
registers. Device 1 error flags are read/reset from the SSTS register.
0 = Software clears this bit by writing a 1 to it.
13
Received Master Abort Status (RMAS1). (Not Applicable). Hardwired to 0.
12
Received Target Abort Status (RTAS1). (Not Applicable). Hardwired to 0.
11
Signaled Target Abort Status (STAS1). (Not Applicable). Hardwired to 0.
10:9
DEVSEL# Timing (DEVT1). (Not Applicable). Hardwired to 00b.
8
Data Parity Detected (DPD1). (Not Applicable). Hardwired to 0.
7
Fast Back-to-Back (FB2B1). (Not Applicable). Hardwired to 0.
6
Reserved.
5
66/60 MHz Capability. (Not Applicable). Hardwired to 1.
4:0
Datasheet
Reserved.
87
82815 GMCH
R
3.5.5.
RID1—Revision Identification Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
08h
02h (see Spec. Update document for latest information.)
Read-Only
8 bits
This register contains the revision number of the GMCH Device 1. These bits are read-only and writes to
this register have no effect.
Bit
7:0
Description
Revision Identification Number. This 8-bit value indicates the revision identification number for the
GMCH Device 1.
02h = A-2 Stepping
3.5.6.
SUBC1—Sub-Class Code Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
0Ah
04h
Read-Only
8 bits
This register contains the Sub-Class Code for the GMCH Device 1. This code is 04h indicating a
PCI-PCI Bridge device. The register is read-only.
Bit
7:0
Description
Sub-Class Code (SUBC1). This 8-bit value indicates the category of Bridge of the GMCH.
04h = Host Bridge.
3.5.7.
BCC1—Base Class Code Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
0Bh
06h
Read-Only
8 bits
This register contains the Base Class Code of the GMCH Device 1. This code is 06h indicating a Bridge
device. This register is read-only.
Bit
7:0
Description
Base Class Code (BASEC). This 8-bit value indicates the Base Class Code for the GMCH Device 1.
06h = Bridge device.
88
Datasheet
82815 GMCH
R
3.5.8.
MLT1—Master Latency Timer Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
0Dh
00h
Read/Write
8 bits
This functionality is not applicable. It is described here since these bits should be implemented as a
read/write to prevent standard PCI-PCI bridge configuration software from getting “confused”.
Bit
3.5.9.
Description
7:3
Not applicable but supports read/write operations. (Reads return previously written data.)
2:0
Reserved.
HDR1—Header Type Register (Device 1)
Address Offset:
Default:
Access:
Size:
0Eh
01h
Read-Only
8 bits
This register identifies the header layout of the configuration space. No physical register exists at this
location.
Bit
7:0
3.5.10.
Descriptions
This read-only field always returns 01h when read. Writes have no effect.
PBUSN—Primary Bus Number Register (Device 1)
Address Offset:
Default:
Access:
Size:
18h
00h
Read-Only
8 bits
This register identifies that the “virtual” PCI-PCI bridge is connected to bus #0.
Bit
7:0
Datasheet
Descriptions
Bus Number. Hardwired to 0.
89
82815 GMCH
R
3.5.11.
SBUSN—Secondary Bus Number Register (Device 1)
Address Offset:
Default:
Access:
Size:
19h
00h
Read /Write
8 bits
This register identifies the bus number assigned to the second bus side of the “virtual” PCI-PCI bridge
(i.e., to PCI1/AGP). This number is programmed by the PCI configuration software to allow mapping of
configuration cycles to PCI1/AGP.
Bit
7:0
3.5.12.
Descriptions
Bus Number. Programmable
SUBUSN—Subordinate Bus Number Register (Device 1)
Address Offset:
Default:
Access:
Size:
1Ah
00h
Read /Write
8 bits
This register identifies the subordinate bus (if any) that resides at the level below PCI1/AGP. This
number is programmed by the PCI configuration software to allow mapping of configuration cycles to
PCI1/AGP.
Bit
7:0
90
Descriptions
Bus Number. Programmable
Datasheet
82815 GMCH
R
3.5.13.
SMLT—Secondary Master Latency Timer Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
1Bh
00h
Read/Write
8 bits
This register controls the bus tenure of the GMCH on AGP/PCI. SMLT is an 8-bit register that controls
the amount of time the GMCH, as a AGP/PCI bus master, can burst data on the AGP/PCI Bus. The count
value is an 8-bit quantity; however, SMLT[2:0] are reserved and assumed to be 0 when determining the
count value. The GMCH’s SMLT is used to guarantee to the AGP master a minimum amount of the
system resources. When the GMCH begins the first PCI bus cycle after being granted the bus, the counter
is loaded and enabled to count from the assertion of FRAME#. If the count expires while the GMCH’s
grant is removed (due to AGP master request), the GMCH will lose the use of the bus and the AGP
master agent may be granted the bus. If the GMCH’s bus grant is not removed, the GMCH continues to
own the AGP/PCI bus, regardless of the SMLT expiration or idle condition. Note that the GMCH must
always properly terminate a AGP/PCI transaction, with FRAME# negation prior to the final data transfer.
The number of clocks programmed in the SMLT represents the guaranteed time slice (measured in
66 MHz PCI clocks) allotted to the GMCH, after which it must complete the current data transfer phase
and then surrender the bus as soon as its bus grant is removed. For example, if the SMLT is programmed
to 18h, the value is 24 AGP clocks. The default value of SMLT is 00h and disables this function. When
the SMLT is disabled, the burst time for the GMCH is unlimited (i.e., the GMCH can burst forever).
7
3
2
Secondary MLT Counter Value
Bit
Datasheet
0
Reserved
Description
7:3
Secondary MLT Counter Value. Default=0 (i.e., SMLT disabled)
2:0
Reserved.
91
82815 GMCH
R
3.5.14.
IOBASE—I/O Base Address Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
1Ch
F0h
Read/Write
8 bits
This register control the processor to PCI1/AGP I/O access routing based on the following formula:
IO_BASE ≤ address ≤ IO_LIMIT
Only the upper 4 bits are programmable. For the purpose of address decode, address bits A[11:0] are
treated as 0. Thus, the bottom of the defined I/O address range will be aligned to a 4-KB boundary.
Note:
BIOS must not set this register to 00h; otherwise, 0CF8h/0CFCh accesses will be forwarded to AGP.
7
4
3
I/O Address Base
Bit
92
0
I/O Addressing Capability
Description
7:4
I/O Address Base. Corresponds to A[15:12] of the I/O address. (Default=Fh)
3:0
I/O Addressing Capability. Hardwired to 0h indicating that only 16 bit I/O addressing is supported. Bits
[31:16] of the I/O base address are assumed to be 0000h.
Datasheet
82815 GMCH
R
3.5.15.
IOLIMIT—I/O Limit Address Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
1Dh
00h
Read/Write
8 bits
This register controls the processor to PCI1/AGP I/O access routing based on the following formula:
IO_BASE ≤ address ≤ IO_LIMIT
Only upper 4 bits are programmable. For the purpose of address decode address bits A[11:0] are
assumed to be FFFh. Thus, the top of the defined I/O address range will be at the top of a 4-KB aligned
address block.
7
4
3
I/O Address Limit
Bit
Datasheet
0
Reserved
Description
7:4
I/O Address Limit. Corresponds to A[15:12] of the I/O address. (Default=0)
3:0
Reserved. (Only 16 bit addressing supported.)
93
82815 GMCH
R
3.5.16.
SSTS—Secondary PCI-PCI Status Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
1E–1Fh
02A0h
Read-Only, Read/Write Clear
16 bits
SSTS is a 16-bit status register that reports the occurrence of error conditions associated with the
secondary side (i.e., PCI1/AGP side ) of the “virtual” PCI-PCI bridge embedded within GMCH.
15
14
13
12
11
Det. Parity
Error
Rec Sys
Error
(HW=0)
Rec
Master
Abort
Rec
Target
Abort
Sig Target
Abort
(HW=0)
7
6
5
FB2B
(HW=1)
Reserved
66/60 MHz
Cap
(HW=1)
10
9
DEVSEL Timing
(HW=01b; medium)
8
Data
Parity Det.
(HW=0)
4
0
Reserved
Bit
Descriptions
15
Detected Parity Error (DPE1). Note that the function of this bit is not affected by the PERRE1 bit. Also
note that PERR# is not implemented in the GMCH.
1 = GMCH detected a parity error in the address or data phase of PCI1/AGP bus transactions.
0 = Software sets DPE1 to 0 by writing a 1 to this bit.
14
13
Received System Error (SSE1). Hardwired to 0. The GMCH does not have an SERR# signal pin.
Received Master Abort Status (RMAS1).
1 = GMCH terminated a Host-to-PCI1/AGP with an unexpected master abort.
0 = Software resets this bit to 0 by writing a 1 to it.
12
Received Target Abort Status (RTAS1).
1 = GMCH-initiated transaction on PCI1/AGP is terminated with a target abort.
0 = Software resets RTAS1 to 0 by writing a 1 to it.
11
Signaled Target Abort Status (STAS1). Hardwired to 0. The GMCH does not generate target abort on
PCI1/AGP.
10:9
DEVSEL# Timing (DEVT1). This 2-bit field indicates the timing of the DEVSEL# signal when the
GMCH responds as a target on PCI1/AGP, and is hard-wired to the value 01b (medium) to indicate the
time when a valid DEVSEL# can be sampled by the initiator of the PCI cycle.
8
Data Parity Detected (DPD1). Hardwired to 0. GMCH does not implement G_PERR# function.
However, data parity errors are still detected and reported using SERR hub interface special cycles (if
enabled by SERRE1 and the BCTRL register, bit 0).
7
Fast Back-to-Back (FB2B1). Hardwired to 1. The GMCH as a target supports fast back-to-back
transactions on PCI1/AGP.
6
Reserved.
5
66/60 MHz Capability. Hardwired to 1.
4:0
94
Reserved.
Datasheet
82815 GMCH
R
3.5.17.
MBASE—Memory Base Address Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
20–21h
FFF0h
Read/Write
16 bits
This register controls the processor to PCI1 non-prefetchable memory access routing based on the
following formula:
MEMORY_BASE ≤ address ≤ MEMORY_LIMIT
The upper 12 bits of the register are read/write and correspond to the upper 12 address bits A[31:20] of
the 32 bit address. The bottom 4 bits of this register are read-only and return zeroes when read. The
configuration software must initialize this register. For the purpose of address decode, address bits
A[19:0] are assumed to be 0. Thus, the bottom of the defined memory address range will be aligned to a
1 MB boundary.
15
4
Memory Address Base
Bit
15: 4
3:0
Datasheet
3
0
Reserved
Description
Memory Address Base (MEM_BASE). Corresponds to A[31:20] of the memory address.
Reserved.
95
82815 GMCH
R
3.5.18.
MLIMIT—Memory Limit Address Register (Device 1)
Address Offset:
Default Value:
Access:
Size:
22–23h
0000h
Read/Write
16 bits
This register controls the processor to PCI1 non-prefetchable memory access routing based on the
following formula:
MEMORY_BASE ≤ address ≤ MEMORY_LIMIT
The upper 12 bits of the register are read/write and correspond to the upper 12 address bits A[31:20] of
the 32 bit address. The bottom 4 bits of this register are read-only and return zeroes when read. The
configuration software must initialize this register. For the purpose of address decode, address bits
A[19:0] are assumed to be FFFFFh. Thus, the top of the defined memory address range will be at the top
of a 1 MB aligned memory block.
15
4
Memory Address Limit
Bit
96
3
0
Reserved
Description
15: 4
Memory Address Limit (MEM_LIMIT). Corresponds to A[31:20] of the memory address. (Default=0)
3:0
Reserved.
Note:
Memory range covered by MBASE and MLIMIT registers are used to map non-prefetchable
PCI1/AGPaddress ranges (typically, where control/status memory-mapped I/O data structures of the
graphics controller will reside) and PMBASE and PMLIMIT are used to map prefetchable address
ranges (typically, graphics local memory). This segregation allows application of USWC space attribute
to be performed in a true plug-and-play manner to the prefetchable address range for improved processor
–AGP memory access performance.
Note:
Configuration software is responsible for programming all address range registers (prefetchable, nonprefetchable) with the values that provide exclusive address ranges (i.e., prevent overlap with each other
and/or with the ranges covered with the main memory). There is no provision in the GMCH hardware to
enforce prevention of overlap and operations of the system in the case of overlap are not guaranteed.
Datasheet
82815 GMCH
R
3.5.19.
PMBASE—Prefetchable Memory Base Address Register
(Device 1)
Address Offset:
Default Value:
Access:
Size:
24–25h
FFF0h
Read/Write
16 bits
This register controls the processor to PCI1 prefetchable memory accesses routing based on the
following formula:
PREFETCHABLE_MEMORY_BASE ≤ address ≤ PREFETCHABLE_MEMORY_LIMIT
The upper 12 bits of the register are read/write and correspond to the upper 12 address bits A[31:20] of
the 32 bit address. The bottom 4 bits of this register are read-only and return zeroes when read. The
configuration software must initialize this register. For the purpose of address decode, address bits
A[19:0] are assumed to be 0. Thus, the bottom of the defined memory address range will be aligned to a
1 MB boundary.
15
4
Prefetchable Memory Address Base
Bit
15: 4
3:0
Datasheet
3
0
Reserved
Description
Prefetchable Memory Address Base(PMEM_BASE). Corresponds to A[31:20] of the memory
address. (Default=FFFh)
Reserved.
97
82815 GMCH
R
3.5.20.
PMLIMIT—Prefetchable Memory Limit Address Register
(Device 1)
Address Offset:
Default Value:
Access:
Size:
26–27h
0000h
Read/Write
16 bits
This register controls the processor to PCI1 prefetchable memory accesses routing based on the
following formula.
PREFETCHABLE_MEMORY_BASE ≤ address ≤ PREFETCHABLE_MEMORY_LIMIT
The upper 12 bits of the register are read/write and correspond to the upper 12 address bits A[31:20] of
the 32 bit address. The bottom 4 bits of this register are read-only and return zeroes when read. The
configuration software must initialize this register. For the purpose of address decode, address bits
A[19:0] are assumed to be FFFFFh. Thus, the top of the defined memory address range will be at the top
of a 1 MB aligned memory block.
15
4
Prefetchable Memory Address Limit
Bit
15: 4
3:0
Note:
98
3
0
Reserved
Description
Prefetchable Memory Address Limit (PMEM_LIMIT). Corresponds to A[31:20] of the memory
address. (Default=0)
Reserved.
Prefetchable memory range is supported to allow segregation by the configuration software between the
memory ranges that must be defined as UC and the ones that can be designated as a USWC
(i.e., prefetchable) from the processor perspective.
Datasheet
82815 GMCH
R
3.5.21.
BCTRL—PCI-PCI Bridge Control Register (Device 1)
Address Offset:
Default:
Access:
Size
3Eh
00h
Read/Write
8 bits
This register provides extensions to the PCICMD1 register that are specific to PCI-PCI bridges. The
BCTRL provides additional control for the secondary interface (i.e., PCI1/AGP) as well as some bits that
affect the overall behavior of the “virtual” PCI-PCI bridge embedded in GMCH (e.g., VGA compatible
address ranges mapping).
7
6
5
4
3
FB2B EN
Sec Bus
Reset
Reserved
Reserved
VGA EN
Bit
Datasheet
2
1
0
SERR#
EN
Parity Err
Response
EN
Description
7
Fast Back to Back Enable. Hardwired to 0. Since there is only one target allowed on AGP, this bit is
meaningless.
6
Secondary Bus Reset. Hardwired to 0. The GMCH does not support generation of reset via this bit on
the AGP. Note that the only way to perform a hard reset of the AGP is via the system reset, either
initiated by software or hardware via the I/O Controller Hub.
5
Master Abort Mode. Hardwired to 0. This means that when acting as a master on AGP/PCI1, the
GMCH will drop writes on the “floor” and return all 1s during reads when a Master Abort occurs.
4
Reserved.
99
82815 GMCH
R
Bit
Description
3
VGA Enable. This bit works with the MDA present bit (GMCHCFG[3]) of device 0 to control the routing
of processor-initiated transactions targeting VGA compatible I/O and memory address ranges. When
this bit is set, the GMCH forwards the following processor accesses to the AGP:
• Memory accesses in the range 0A0000h to 0BFFFFh
• I/O addresses where A[9:0] are in the ranges 3B0h to 3BBh and 3C0h to 3DFh
(inclusive of ISA address aliases – A[15:10] are not decoded)
1 = Enable. Forwarding of these accesses issued by the processor is independent of the I/O address
and memory address ranges defined by the previously defined Base and Limit registers.
Forwarding of these accesses is also independent of the settings of bit 2 (ISA Enable) of this
register if this bit is 1. If the VGA enable bit is set, accesses to IO address range x3BCh–x3BFh are
forwarded to the hub interface. If the VGA enable bit is not set, accesses to IO address range
x3BCh–x3BFh are treated just like any other IO accesses (i.e., cycles are forwarded to AGP, if the
address is within IOBASE and IOLIMIT and ISA enable bit is not set; otherwise, they are forwarded
to hub interface).
0 = Disable (default). VGA compatible memory and I/O range accesses are not forwarded to AGP;
rather, they are mapped to primary PCI unless they are mapped to AGP via I/O and memory range
registers defined above (IOBASE, IOLIMIT, MBASE, MLIMIT, PMBASE, PMLIMIT).
The following table shows the behavior for all combinations of MDA and VGA:
VGA
2
MDA
Behavior
0
0
0
1
All references to MDA and VGA Go To hub interface
Illegal combination (DO NOT USE)
1
0
All references To VGA Go To AGP MDA-only references (I/O Address 3BF
and aliases) will go to hub interface.
1
1
VGA references go to AGP/PCI; MDA references go to the hub interface
ISA Enable. Modifies the response by the GMCH to an I/O access issued by the processor that targets
ISA I/O addresses. This applies only to I/O addresses that are enabled by the IOBASE and IOLIMIT
registers.
1 = Enable. GMCH will not forward to PCI1/AGP any I/O transactions addressing the last 768 bytes in
each 1 KB block, even if the addresses are within the range defined by the IOBASE and IOLIMIT
registers. Instead of going to PCI1/AGP, these cycles are forwarded to the hub interface where
they can eventually be subtractively or positively claimed by the ISA bridge.
0 = Disable (default). All addresses defined by the IOBASE and IOLIMIT for processor I/O transactions
are mapped to PCI1/AGP.
1
SERR# Enable. Hardwired to 0. This bit normally controls forwarding SERR# on the secondary
interface to the primary interface. The GMCH does not support the SERR# signal on the AGP/PCI1
bus.
0
Parity Error Response Enable. Controls GMCH’s response to data phase parity errors on PCI1/AGP.
G_PERR# is not implemented by the GMCH.
1 = Enable. Address and data parity errors on PCI1 are reported via SERR messaging, if enabled by
SERRE1.
0 = Disable. Address and data parity errors on PCI1/AGP are not reported via SERR messaging. Other
types of error conditions can still be signaled via SERR messaging independent of this bit’s state.
100
Datasheet
82815 GMCH
R
3.5.22.
ERRCMD1—Error Command Register (Device 1)
Address Offset:
Default:
Access:
Size
40h
00h
Read/Write
8 bits
7
1
Reserved
Bit
7:1
0
0
SERR on
Rec Targ.
Abort on
AGP/PCI
Descriptions
Reserved.
SERR on Receiving Target Abort on AGP/PCI.
1 = Enable. The GMCH generates an SERR hub interface special cycle when an GMCH-originated
AGP/PCI cycle is terminated with a Target Abort.
0 = Disable. Reporting of this condition is disabled.
Datasheet
101
82815 GMCH
R
3.6.
Graphics Device Registers
(Device 2: VISIBLE IN GFX Mode Only)
These registers are accessible through the configuration mechanism defined in an earlier section of this
document.
Table 8. Device 2 Configuration Space Address Map (Internal Graphics)
Address
Offset
Register
Symbol
Register Name
Default Value
Access
Type
00–01h
VID2
Vendor Identification
8086h
RO
02–03h
DID2
Device Identification
1132h
RO
04–05h
PCICMD2
PCI Command
0004h
R/W
06–07h
PCISTS2
PCI Status
02B0h
RO
08h
RID2
Revision Identification
02h (see note)
RO
09h
PI
Programming Interface
00h
RO
0Ah
SUBC2
Sub-Class Code
00h
RO
0Bh
BCC2
Base Class Code
03h
RO
0Ch
CLS
Cache Line Size
00h
RO
0Dh
MLT2
Master Latency Timer
00h
RO
0Eh
HDR2
Header Type
01h
RO
0Fh
BIST
BIST Register
00h
RO
10–13h
GMADR
Graphics Memory Range Address
00000008h
R/W
14–17h
MMADR
Memory Mapped Range Address
00000000h
R/W
18–2Bh

00h

2C–2Dh
SVID
2E–2Fh
SID
30–33h
ROMADR
34h
CAPPOINT
35–3Bh

3Ch
Reserved
Subsystem Vendor ID
0000h
R/WO
Subsystem ID
0000h
R/WO
00000000h
RO
Capabilities Pointer
DCh
RO
Reserved
00h

INTRLINE
Interrupt Line
00h
R/W
3Dh
INTRPIN
Interrupt Pin
01h
RO
3Eh
MINGNT
Minimum Grant
00h
RO
3Fh
MAXLAT
Maximum Latency
00h
RO
40–DBh

Reserved
00h

DC–DDh
PM_CAPID
Power Management Capabilities
0001h
RO
DE–DFh
PM_CAP
Power Management Capabilities
0022h
RO
E0–E1h
PM_CS
Power Management Control
0000h
R/W
E2–FFh

00h

Video Bios ROM Base Address
Reserved
Note: See Specification Update document for latest information.
102
Datasheet
82815 GMCH
R
3.6.1.
VID2—Vendor Identification Register (Device 2)
Address Offset:
Default Value:
Attribute:
Size:
00h−01h
8086h
Read-Only
16 bits
The VID Register contains the vendor identification number. This 16-bit register combined with the
Device Identification Register uniquely identify any PCI device. Writes to this register have no effect.
Bit
15:0
3.6.2.
Description
Vendor Identification Number. This 16-bit value is assigned to Intel.
DID2—Device Identification Register (Device 2)
Address Offset:
Default Value:
Attribute:
Size:
02h−03h
1132h
Read-Only
16 bits
This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device.
Writes to this register have no effect.
Bit
15:0
Description
Device Identification Number. This 16 bit value is assigned to the internal graphics device of the
GMCH.
1132h = Device ID for Device 2.
Datasheet
103
82815 GMCH
R
3.6.3.
PCICMD2—PCI Command Register (Device 2)
Address Offset:
Default:
Access:
Size:
04h−05h
0004h
Read-Only, Read/Write
16 bits
This 16-bit register provides basic control over the chipset’s ability to respond to PCI cycles. The
PCICMD Register in the GMCH disables PCI compliant master accesses to main memory.
15
10
Reserved (0)
9
8
FB2B
(Not Impl)
SERR En
(Not Impl)
7
6
5
4
3
2
1
0
Addr/Data
Stepping
(Not Impl)
Parity
Error En
(Not Impl)
VGA Pal
Sn
(Not Impl)
Mem WR
& Inval En
(Not Impl)
Special
Cycle En
(Not Impl)
Bus
Master En
(Enabled)
Mem Acc
En
I/O Acc En
Bits
15:10
Description
Reserved.
9
Fast Back-to-Back (FB2B)
RO. (Not Implemented). Hardwired to 0.
8
SERR# Enable (SERRE) RO. (Not Implemented). Hardwired to 0.
7
Address/Data Stepping
RO. (Not Implemented). Hardwired to 0.
6
Parity Error Enable (PERRE) RO. (Not Implemented). Hardwired to 0. Since the GMCH belongs to
the category of devices that does not corrupt programs or data in system memory or hard drives, the
GMCH ignores any parity error that it detects and continues with normal operation.
5
Video Palette Snooping (VPS) RO. Hardwired to 0. Disables snooping.
4
Memory Write and Invalidate Enable (MWIE) RO. Hardwired to 0. GMCH does not support
memory write and invalidate commands.
3
Special Cycle Enable (SCE) RO. Hardwired to 0. GMCH ignores Special cycles.
2
Bus Master Enable (BME) RO. Hardwired to 1 to enable GMCH to function as a PCI compliant
master.
1
Memory Access Enable (MAE) R/W. This bit controls the GMCH’s response to memory space
accesses.
0 = Disable (default).
1 = Enable.
0
I/O Access Enable (IOAE) R/W. This bit controls the GMCH’s response to I/O space accesses.
0 = Disable (default).
1 = Enable.
104
Datasheet
82815 GMCH
R
3.6.4.
PCISTS2—PCI Status Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
06h−07h
02B0h
Read-Only
16 bits
PCISTS2 reports the occurrence of a PCI compliant master abort and PCI compliant target abort.
PCISTS2 also indicates the DEVSEL# timing that has been set by the GMCH hardware.
15
14
13
12
11
Detected
Par Error
(HW=0)
Sig Sys
Error
(HW=0)
Recog
Mast Abort
Sta
(HW=0)
Rec
Target
Abort Sta
(HW=0)
Sig Target
Abort Sta
(HW=0)
7
6
5
4
FB2B
(HW=1)
User Def
Format
(HW=0)
66 MHz
PCI Cap
(HW=1)
Cap List
(HW=1)
Bits
10
9
DEVSEL# Timing
(HW=01)
8
Data Par
Detected
(HW=0)
3
0
Reserved
Description
15
Detected Parity Error (DPE) RO. Hardwired to 0. The chipset does not detect parity.
14
Signaled System Error (SSE) RO. Hardwired to 0. The chipset’s graphics device never asserts
SERR#.
13
Received Master Abort Status (RMAS) RO. Hardwired to 0. The chipset’s graphics device never
gets a Master Abort.
12
Received Target Abort Status (RTAS) RO.. Hardwired to 0. The chipset’s graphics device never
gets a Target Abort.
11
Signaled Target Abort Status (STAS). Hardwired to 0. The chipset does not use target abort
semantics.
10:9
DEVSEL# Timing (DEVT) RO. This 2-bit field indicates the timing of the DEVSEL# signal when
GMCH responds as a target.
01 = Medium decode device (hardwired).
8
Data Parity Detected (DPD) RO. Hardwired to 0. Since Parity Error Response is hardwired to
disabled (and GMCH does not do any parity detection), this bit is not used.
7
Fast Back-to-Back (FB2B). Hardwired to 1. The chipset accepts fast back-to-back when the
transactions are not to the same agent.
6
User Defined Format (UDF). Hardwired to 0.
5
66 MHz PCI Capable (66C). Hardwired to 1. This indicates that the chipset is 66 MHz PCI capable.
4
CAP LIST
RO. This bit is set to 1 to indicate that the register at 34h provides an offset into the
function’s PCI Configuration Space containing a pointer to the location of the first item in the list.
3:0
Datasheet
Reserved.
105
82815 GMCH
R
3.6.5.
RID2—Revision Identification Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
08h
02h (see Spec. Update document for latest information.)
Read-Only
8 bits
This register contains the revision number of the chipset. These bits are read-only and writes to this
register have no effect.
Bits
7:0
3.6.6.
Description
Revision Identification Number. This is an 8-bit value that indicates the revision identification
number for the GMCH. The four MSBs are for process differentiation and the four LSBs indicate
stepping.
PI—Programming Interface Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
09h
00h
Read-Only
8 bits
This register contains the device programming interface for the GMCH.
Bits
7:0
Description
Programming Interface (PI). Hardwired to 00h.
00h = Display controller.
3.6.7.
SUBC2—Sub-Class Code Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Ah
00h
Read-Only
8 bits
This register contains the Sub-Class Code for the GMCH Device 2.
Bit
7:0
Description
Sub-Class Code (SUBC). This is an 8-bit value that indicates the category of display controller of the
GMCH.
00h = VGA compatible device.
106
Datasheet
82815 GMCH
R
3.6.8.
BCC2—Base Class Code Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Bh
03h
Read-Only
8 bits
This register contains the Base Class Code of the GMCH Function #1.
Bit
7:0
Description
Base Class Code (BASEC). This is an 8-bit value that indicates the Base Class Code for GMCH.
03h = Display controller.
3.6.9.
CLS—Cache Line Size Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Ch
00h
Read-Only
8 bits
The GMCH does not support this register as a PCI slave.
Bits
7:0
3.6.10.
Description
Cache Line Size (CLS). This field is hardwired to 0s. The GMCH, as a PCI compliant master, does
not use the Memory Write and Invalidate command and, in general, does not perform operations
based on cache line size.
MLT2—Master Latency Timer Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Dh
00h
Read-Only
8 bits
The GMCH does not support the programmability of the master latency timer because it does not
perform bursts.
Bits
7:0
Datasheet
Description
Master Latency Timer Count Value. Hardwired to 0s.
107
82815 GMCH
R
3.6.11.
HDR2—Header Type Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Eh
00h
Read-Only
8 bits
This register contains the Header Type of the GMCH .
Bits
Description
Header Type (HTYPE). This is an 8-bit value that indicates the Header Type for the chipset.
7:0
00h = Indicating a basic (i.e., single function) configuration space format.
3.6.12.
BIST—BIST Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
0Fh
00h
Read-Only
8 bits
This register is used for control and status of Built In Self Test (BIST).
7
6
BIST
Supported
(HW=0)
Bits
7
6:0
108
0
Reserved
Description
BIST Supported. BIST is not supported. This bit is hardwired to 0.
Reserved.
Datasheet
82815 GMCH
R
3.6.13.
GMADR—Graphics Memory Range Address Register
(Device 2)
Address Offset:
Default Value:
Access:
Size:
10−13h
00000008h
Read/Write, Read-Only
32 bits
This register requests allocation for the GMCH graphics memory. The allocation is for either 32 MB or
64 MB of memory space (selected by bit 0 of the Device 0 MISCC Register) and the base address is
defined by bits [31:25,24].
31
26
Memory Base Address
25
24
16
64 MB
Addr. Mask
15
4
Address Mask (cont)
(HW=0; 32MB addr range)
Address Mask
(HW=0; 32MB addr range)
3
2
Prefetch
Mem En
(HW=1)
1
Memory Type
(HW=0; 32MB addr)
0
Mem/IO
Space
(HW=0)
Bit
Description
31:26
Memory Base Address
R/W. Set by the operating system, these bits correspond to address signals
[31:26].
25
64 MB Address Mask
RO , R/W.
64 MB = If Device 0 MISCC Reg bit 0 = 0, then this bit is read-only with a value of 0, indicating a
memory range of 64 MB.
32 MB = If Device 0 MISCC Reg bit 0 = 1, this bit is R/W, indicating a memory range of 32 MB.
24:4
3
2:1
0
Datasheet
Address Mask
RO. Hardwired to 0s to indicate 32 MB address range.
Prefetchable Memory
RO. Hardwired to 1 to enable prefetching.
Memory Type
RO. Hardwired to 0 to indicate 32-bit address.
Memory/IO Space
RO. Hardwired to 0 to indicate memory space.
109
82815 GMCH
R
3.6.14.
MMADR—Memory Mapped Range Address Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
14−17h
00000000h
Read/Write, Read-Only
32 bits
This register requests allocation for the GMCH registers and instruction ports. The allocation is for
512 KB and the base address is defined by bits [31:19].
31
19
Memory Base Address
(addr bits [31:19])
15
4
Address Mask (cont)
(HW=0; 512 KB addr range)
Address Mask (HW=0;
512 KB addr range)
3
Prefetch
Mem En
(HW=0)
2
1
Memory Type
(HW=0; 32 Mb addr)
0
Mem/IO
Space
(HW=0)
Description
31:19
Memory Base Address
R/W. Set by the operating system, these bits correspond to address signals
[31:19].
18:4
Address Mask
RO. Hardwired to 0s to indicate 512 KB address range.
2:1
0
Prefetchable Memory
RO. Hardwired to 0 to prevent prefetching.
Memory Type
RO. Hardwired to 0s to indicate 32-bit address.
Memory / IO Space
RO. Hardwired to 0 to indicate memory space.
SVID—Subsystem Vendor Identification Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
Bit
15:0
110
16
Bit
3
3.6.15.
18
2C−2Dh
0000h
Read/Write-Once
16 bits
Description
Subsystem Vendor ID—R/WO. This value is used to identify the vendor of the subsystem. The
default value is 0000h. This field should be programmed by BIOS during boot-up. Once written, this
register becomes read-only. This Register can only be cleared by a Reset.
Datasheet
82815 GMCH
R
3.6.16.
SID—Subsystem Identification Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
3.6.17.
2E−2Fh
0000h
Read/Write-Once
16 bits
Bit
Description
15:0
Subsystem ID—R/WO. This value is used to identify a particular subsystem. The default value is
0000h. This field should be programmed by BIOS during boot-up. Once written, this register becomes
Read only. This Register can only be cleared by a Reset.
ROMADR—Video BIOS ROM Base Address Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
30−33h
00000000h
Read Only
32 bits
The GMCH does not use a separate BIOS ROM; therefore, this is hardwired to 0s.
31
18
ROM Base Address
(addr bits [31:18])
15
11
10
1
Reserved
Bit
3.6.18.
16
Address Mask (HW=0;
256 KB addr range)
Address Mask (cont)
(HW=0; 256 KB addr range)
0
ROM
BIOS En
(HW=0)
Description
31:18
ROM Base Address
RO. Hardwired to 0s.
17:11
Address Mask
RO. Hardwired to 0s to indicate 256 KB address range.
10:1
Reserved. Hardwired to 0s.
0
17
ROM BIOS Enable
RO. 0 = ROM not accessible.
CAPPOINT—Capabilities Pointer Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
34h
DCh
Read Only
8 bits
Bit
Description
7:0
Pointer to the start of AGP standard register block. Since there is no AGP bus on the GMCH , the
value of this field is DCh.
DCh = Points to the Power Management Capabilities ID Register
Datasheet
111
82815 GMCH
R
3.6.19.
INTRLINE—Interrupt Line Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
Bit
7:0
3.6.20.
3Ch
00h
Read/Write
8 bits
Descriptions
Interrupt Connection. Used to communicate interrupt line routing information. POST software writes
the routing information into this register as it initializes and configures the system. The value in this
register indicates which input of the system interrupt controller that the device’s interrupt pin is
connected.
INTRPIN—Interrupt Pin Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
Bit
7:0
3Dh
01h
Read Only
8 bits
Descriptions
Interrupt Pin. As a single function device, the GMCH specifies INTA# as its interrupt pin.
01h = INTA#.
3.6.21.
MINGNT—Minimum Grant Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
Bit
7:0
3.6.22.
Descriptions
Minimum Grant Value. The GMCH does not burst as a PCI compliant master. (Default=00h).
MAXLAT—Maximum Latency Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
Bit
7:0
112
3Eh
00h
Read Only
8 bits
3Fh
00h
Read Only
8 bits
Descriptions
Maximum Latency Value. The GMCH has no specific requirements for how often it needs to access
the PCI bus. (Default=00h).
Datasheet
82815 GMCH
R
3.6.23.
PM_CAPID—Power Management Capabilities ID Register
(Device 2)
Address Offset:
Default Value:
Access:
Size:
DCh−DDh
0001h
Read Only
16 bits
15
8
7
0
NEXT_PTR
CAP_ID
Bits
3.6.24.
Description
15:8
Next Pointer (NEXT_PTR). This contains a pointer to next item in the capabilities list. This the final
capability in the list and must be set to 00h.
7:0
Capability Identificaiton (CAP_ID). SIG defines this ID is 01h for power management.
PM_CAP—Power Management Capabilities Register (Device 2)
Address Offset:
Default Value:
Access:
Size:
DEh−DFh
0022h
Read Only
16 bits
15
11
PME Support (HW=0)
7
6
Reserved
4
3
Dev
Specific
Init
(HW=1)
Aux Pwr
Src
(HW=0)
PME
Clock
(HW=0)
9
8
D2
(HW=0)
D1
(HW=0)
Reserved
2
0
Version
Bits
Description
15:11
PME Support. This field indicates the power states in which the GMCH may assert PME#. Hardwired
to 0 to indicate that the GMCH does not assert the PME# signal.
10
D2. Hardwired to 0 to indicate D2 power management state is not supported.
9
D1. Hardwired to0 to indicate that D1 power management state is NOT supported.
8:6
Reserved. Read as 0s.
5
Device Specific Initialization (DSI). Hardwired to 1 to indicate that special initialization of the GMCH
is required before generic class device driver is to use it.
4
Auxiliary Power Source. Hardwired to 0.
3
PME Clock. Hardwired to 0 to indicate the GMCH does not support PME# generation.
2:0
Datasheet
5
10
Version. Hardwired to 010b to indicate there are 4 bytes of power management registers
implemented and that this device complies with revision 1.1 of the PCI Power Management Interface
Specification.
113
82815 GMCH
R
3.6.25.
PM_CS—Power Management Control/Status Register
(Device 2)
Address Offset:
Default Value:
Access:
Size:
15
PME Sta
(HW=0)
14
E0h−E1h
0000h
Read/Write
16 bits
13
12
Data Scale (Reserved)
9
Data_Select (Reserved)
7
2
Reserved
Bits
8
PME En
1
0
PowerState
Description
15
PME_Status RO. This bit is 0 to indicate that the GMCH does not support PME# generation from
D3 (cold).
14:13
Data Scale (Reserved) RO. The GMCH does not support data register. This bit always returns 0
when read; write operations have no effect.
12:9
Data_Select (Reserved) RO. The GMCH does not support data register. This bit always returns 0
when read; write operations have no effect.
8
PME_En
R/W. This bit is 0 to indicate that PME# assertion from D3 (cold) is disabled.
7:2
Reserved. Always returns 0s when read; write operations have no effect.
1:0
PowerState
R/W. This field indicates the current power state of the GMCH and can be used to set
the GMCH into a new power state. If software attempts to write an unsupported state to this field,
write operation must complete normally on the bus; data is discarded and no state change occurs.
00 = D0
01 = D1 (Not supported in the GMCH )
10 = D2 (Not supported in the GMCH )
11 = D3
114
Datasheet
82815 GMCH
R
3.7.
Display Cache Interface
The Display Cache (DC) interface control registers are located in Memory Space. This section describes
the DC interface registers. These registers are accessed using [MMADR+Offset]. These registers are
memory-mapped only.
3.7.1.
DRT—DRAM Row Type
Memory Offset Address:
Default value:
Access:
Size:
3000h
00h
Read / write
8 bit
This 8-bit register identifies whether or not the display cache is populated. Memory mapped only.
7
1
Reserved
Bit
7:1
0
0
DRAM
Populated
Description
Reserved
DRAM Populated (DP). The bit in this register indicates whether or not the display cache is populated.
0 = No Display Cache
1 = 4 MB Display Cache
Datasheet
115
82815 GMCH
R
3.7.2.
DRAMCL—DRAM Control Low
Memory Offset Address:
Default value:
Access:
Size:
3001h
17h
Read / write
8 bit
7
5
Reserved
4
3
2
1
0
Paging
Mode
Control
RAS-toCAS
Override
CAS#
Latency
RAS#
Riming
RAS#
Precharge
Timing
Bit
7:5
4
Description
Reserved
Paging Mode Control (PMC).
0 = Page Open Mode. In this mode the GMCH memory controller tends to leave pages open.
1 = Page Close Mode. In this mode the GMCH memory controller tends to leave pages closed.
3
RAS-to-CAS Override (RCO). In units of display cache clock periods indicates the RAS#-to-CAS# delay
(tRCD). (i.e., row activate command to read/write command)
0 = Determined by CL bit (default)
1=2
2
1
CAS# Latency (CL). In units of local memory clock periods.
Bit
CL
0
2
2
1
3
3 (default)
RAS# Timing (RT). This bit controls RAS# active to precharge, and refresh to RAS# active delay (in
local memory clocks).
Bit
0
RAS#-to-CAS# delay (tRCD)
RAS# act. To precharge (tRAS)
Refresh to RAS# act. (tRC)
0
5
8
1
7
10 (default)
RAS# Precharge Timing (RPT). This bit controls RAS# precharge (in local memory clocks).
0=2
1 = 3 (default)
116
Datasheet
82815 GMCH
R
3.7.3.
DRAMCH—DRAM Control High
Memory Offset Address:
Default value:
Access:
Size:
7
3002h
08h
Read / write
8 bit
5
Reserved
Bit
4
3
DRAM Refresh Rate
2
0
Special Mode Select
Description
7:5
Reserved
4:3
DRAM Refresh Rate (DRR). DRAM refresh is controlled using this field. Disabling refresh results in the
eventual loss of DRAM data; refresh can be briefly disabled without data loss. The field must be set to
normal refresh as soon as possible once DRAM testing is completed.
00 = Refresh Disabled
01 = Refresh Enabled (default)
10 = Reserved
11 = Reserved
2:0
Special Mode Select (SMS). These bits select special SDRAM modes used for testing and
initialization. The NOP command must be programmed first before any other command can be issued.
000 = Normal SDRAM mode (Normal, default).
001 = NOP Command Enable (NCE). This state forces cycles to DRAM to generate SDRAM NOP
commands.
010 = All Banks Precharge Command Enable (ABPCE). This state forces cycles to DRAM to
generate an all banks precharge command.
011 = Mode Register Command Enable (MRCE). This state forces all cycles to DRAM to be converted
into MRS commands. The command is driven on the LMA[11:0] lines. LMA[2:0] correspond to the
burst length, LMA[3] corresponds to the wrap type, and LMA[6:4] correspond to the latency mode.
LMA[11:7] are driven to 00000 by the GMCH,
BIOS must select an appropriate host address for each row of memory such that the right
commands are generated on the LMA[6:0] lines, taking into account the mapping of host
addresses to display cache addresses.
100 = CBR Cycle Enable (CBRCE). This state forces cycles to DRAM to generate SDRAM CBR
refresh cycles.
101 = Reserved.
11X = Reserved.
Datasheet
117
82815 GMCH
R
3.8.
Display Cache Detect and Diagnostic Registers
The following registers are used for display cache detection and diagnostics. These registers can be
accessed via either I/O space or memory space. The memory space addresses listed are offsets from the
base memory address programmed into the MMADR register (Device 2, PCI configuration offset 14h).
For each register, the memory-mapped address offset is the same address value as the I/O address.
3.8.1.
GRX—GRX Graphics Controller Index Register
I/O (and Memory Offset) Address:
Default:
Access:
Size:
7
4
Reserved (0000)
Bit
118
3CEh
0Uh (U=Undefined)
Read/Write
8 bits
3
0
Graphics Controller Register Index
Description
7:4
Reserved. Read as 0s.
3:0
Sequencer Register Index. This field selects any one of the graphics controller registers (GR[00:08]) to
be accessed via the data port at I/O location 3CFh.
Datasheet
82815 GMCH
R
3.8.2.
MSR
Miscellaneous Output
I/O (and Memory Offset) Address:
Default:
Access:
Size:
3C2h  Write;
3CCh Read
00h
See Address above
8 bits
7
2
Reserved
Bit
7:2
1
1
0
A0000h−
BFFFFh
Acc En
Reserved
Descriptions
Reserved
A0000−
−BFFFFh Access Enable. VGA Compatibility bit enables access to the display cache at
A0000h−BFFFFh. When disabled, accesses to system memory are blocked in this region (by not
asserting DEVSEL#). This bit does not block processor access to the video linear frame buffer at other
addresses.
0 = Prevent processor access to the display cache (default).
1 = Allow processor access to display cache.
0
3.8.3.
Reserved
GR06
Miscellaneous Register
I/O (and Memory Offset) Address:
Default:
Access:
Size:
7
3CFh (Index=06h)
0Uh (U=Undefined)
Read/Write
8 bits
4
Reserved
Bit
3
2
Memory Map Mode
1
0
Reserved
Description
7:4
Reserved
3:2
Memory Map Mode. These 2 bits control the mapping of the VGA frame buffer into the processor
address space as follows:
00 = A0000h − BFFFFh
01 = A0000h − AFFFFh
10 = B0000h − B7FFFh
11 = B8000h − BFFFFh
Note: This function is both in standard VGA modes and in extended modes that do not provide linear
frame buffer accesses.
1:0
Datasheet
Reserved
119
82815 GMCH
R
3.8.4.
GR10
Address Mapping
I/O (and Memory Offset) Address:
Default:
Access:
Size:
7
3CFh (Index=10h)
00h
R/W
8 bits
5
Reserved
4
3
2
1
0
Paging to
display
cache
VGA
Buffer
/Memory
Map
Packed
Mode Enbl
Linear
Mapping
Page
Mapping
Bit
7:5
4
Description
Reserved
Page to Display Cache Enable.
0 = Page to VGA Buffer.
1 = Page to Display Cache.
3
VGA Buffer/Memory Map Select.
0 = VGA Buffer (default)
1 = Memory Map.
2
Packed Mode Enable.
0 = Disable (default)
1 = Enable
1
Linear Mapping (PCI).
0 = Disable (default)
1 = Enable
0
Page Mapping Enable. This mode allows the mapping of the vga space allocated in main memory
(non local video memory) mode or all of local memory space through the A0000:AFFFF window which
is a 64 KB page.
0 = Disable (default)
1 = Enable
3.8.5.
GR11
Page Selector
I/O (and Memory Offset) Address:
Default :
Attributes:
Bit
7:0
120
3CFh (Index=11h)
00h
R/W
Description
Page Select. Selects a 64 KB window within the display cache when Page Mapping is enabled to the
display cache.
Datasheet
82815 GMCH
R
4.
Functional Description
This chapter describes the Graphics and Memory Controller Hub (GMCH) interfaces, and boot
sequencing.. The “System Address Map” provides a system-level address memory map and describes the
memory space controls provided by the GMCH.
4.1.
System Address Map
An Intel Pentium III processor, Intel Pentium II processor, or Intel CeleronTM processor system based
on the GMCH, supports 4 GB of addressable memory space and 64 KB+3 of addressable I/O space.
(The P6 bus I/O addressability is 64KB + 3). There is a programmable memory address space under the
1 MB region which can be controlled with programmable attributes of write-only, or read-only. Attribute
programming is described in the Configuration Register Description section. This section focuses on how
the memory space is partitioned and what these separate memory regions are used for. The I/O address
space is explained at the end of this section.
The Intel Pentium III processor, Intel Pentium II processor, and Intel CeleronTM processor supports
addressing of memory ranges larger than 4 GB. The GMCH Host Bridge claims any access over 4 GB by
terminating transaction (without forwarding it to the hub interface). Writes are terminated by dropping
the data and for reads the GMCH returns all zeros on the host bus.
In the following sections, it is assumed that all of the compatibility memory ranges reside on the hub
interface. The exceptions to this rule are the VGA ranges which may be mapped to the internal Graphics
Device.
Note:
Datasheet
The GMCH memory map includes a number of programmable ranges. All of these ranges MUST be
unique and NON-OVERLAPPING. There are NO Hardware Interlocks to prevent problems in the case
of overlapping ranges. Accesses to overlapped ranges may produce indeterminate results.
121
82815 GMCH
R
4.1.1.
Memory Address Ranges
Figure 4 shows a high-level representation of the system memory address map. Figure 5 provides
additional details on mapping specific memory regions as defined and supported by the GMCH.
Figure 4. System Memory Address Map
4 GB
PCI
Memory
Address
Range
Local
Memory
Range
Memory
Mapped
Range
Top of the Main Memory
Main
Memory
Address
Range
0
Independently
Programmable NonOverlapping Memory
Windows
mem_map_s
Figure 5. Detailed Memory System Address Map
System Memory
Space
64 GB
Extended
P6
Memory
4 GB Max TOM
AGP
Graphics
Window Aperture
0FFFFFh
PCI
Memory
Range
1 MB
Upper BIOS Area
(64 KB)
512 MB
0F0000h
0EFFFFh
960 KB
Lower BIOS Area
(64 KB; 16 KB x 4)
0E0000h
0DFFFFh
896 KB
16 MB
Expansion Card
BIOS
and Buffer Area
(128 KB;
16 KB x 8)
Optional ISA Hole
15 MB
Main Memory
Range
1 MB
DOS
Compatibility
Memory
640 KB
0C0000h
0BFFFFh
Optionally
mapped to the
internal AGP
0A0000h
09FFFFh
768 KB
Std PCI/ISA
Video Mem
(SMM Mem)
128 KB
640 KB
DOS Area
(640 KB)
0 MB
000000h
0 KB
mem_map
122
Datasheet
82815 GMCH
R
4.1.2.
Compatibility Area
This area is divided into the following address regions:
• 0–640 KB DOS Area
• 640–768 KB Video Buffer Area
• 768–896 KB in 16KB sections (total of 8 sections) – Expansion Area
• 896–960 KB in 16KB sections (total of 4 sections) – Extended System BIOS Area
• 960 KB–1 MB Memory (BIOS Area) – System BIOS Area
The GMCH supports all sixteen memory segments of interest in the compatibility area. Thirteen of the
memory ranges can be enabled or disabled independently for both read and write cycles.
Table 9. Memory Segments and Their Attributes
Memory Segments
Attributes
Comments
000000h–09FFFFh
Fixed; always mapped to main DRAM
0 to 640K: DOS Region
0A0000h–0BFFFFh
mapped to the hub interface or (AGP or internal
graphics) – configurable as SMM space
Video Buffer (physical DRAM
configurable as SMM space)
0C0000h–0C3FFFh
WE RE
Add-on BIOS
0C4000h–0C7FFFh
WE RE
Add-on BIOS
0C8000h–0CBFFFh
WE RE
Add-on BIOS
0CC000h–0CFFFFh
WE RE
Add-on BIOS
0D0000h–0D3FFFh
WE RE
Add-on BIOS
0D4000h–0D7FFFh
WE RE
Add-on BIOS
0D8000h–0DBFFFh
WE RE
Add-on BIOS
0DC000h–0DFFFFh
WE RE
Add-on BIOS
0E0000h–0E3FFFh
WE RE
BIOS Extension
0E4000h–0E7FFFh
WE RE
BIOS Extension
0E8000h–0EBFFFh
WE RE
BIOS Extension
0EC000h–0EFFFFh
WE RE
BIOS Extension
0F0000h–0FFFFFh
WE RE
BIOS Area
DOS Area (00000h–9FFFh)
The DOS area is 640 KB and is always mapped to the main memory controlled by the GMCH.
Datasheet
123
82815 GMCH
R
Video Buffer Area (A0000h–BFFFFh)
The 128 KB graphics adapter memory region is normally mapped to a legacy video device on the hub
interface/PCI (typically VGA controller). This area is not controlled by attribute bits and processor –
initiated cycles in this region are forwarded to either the hub interface or the AGP/internal graphics
device for termination. This region is also the default region for SMM space.
Accesses to this range are directed to either the hub interface or the AGP/internal graphics device based
on the configuration. The configuration is specified by:
1. AGP on/off configuration bit
2. AGP off: GMS bits of the SMRAM register in the GMCH Device #0 configuration space. There is
additional steering information coming from the Device #2 configuration registers and from some
of the VGA registers in the Graphics device.
3. AGP on: PCICMD1 (PCI-PCI Command) and BCTRL (PCI-PCI Bridge Control) registers in
Device #1 configuration registers
The control is applied for accesses initiated from any of the system interfaces; that is, processor bus, the
hub interface, or AGP(if enabled). Note that for hub interface to AGP/PCI accesses, only memory write
operations are supported. Any AGP/PCI initiated VGA accesses targeting the GMCH will master abort.
For more details, see the descriptions in the configuration registers specified above.
The SMRAM Control register controls how SMM accesses to this space are treated.
Monochrome Adapter (MDA) Range (B0000h–B7FFFh)
Legacy support requires the ability to have a second graphics controller (monochrome) in the system.
In an AGP system, accesses in the standard VGA range are forwarded to the AGP bus (depending on
configuration bits). Since the monochrome adapter may be on the hub interface/PCI (or ISA) bus, the
GMCH must decode cycles in the MDA range and forward them to the hub interface. This capability is
controlled by a configuration bit (MDA bit – Device 0, BEh). In addition to the memory range B0000h
to B7FFFh, the GMCH decodes IO cycles at 3B4h, 3B5h, 3B8h, 3B9h, 3Bah and 3BFh and forwards
them to the hub interface bus
In an internal graphics system, the GMS bits of the SMRAM register in Device #0, bits in the Device 2
PCICMD2 register and bits from some of the VGA registers control this functionality.
Expansion Area (C0000h–DFFFFh)
This 128 KB ISA Expansion region is divided into eight 16 KB segments. Each segment can be assigned
one of four read/write states: read only, write-only, read/write, or disabled. Typically, these blocks are
mapped through GMCH and are subtractively decoded to ISA space. Memory that is disabled is not
remapped.
Extended System BIOS Area (E0000h–EFFFFh)
This 64 KB area is divided into four 16 KB segments. Each segment can be assigned independent read
and write attributes so it can be mapped either to main DRAM or to the hub interface. Typically, this area
is used for RAM or ROM. Memory segments that are disabled are not remapped elsewhere.
124
Datasheet
82815 GMCH
R
System BIOS Area (F0000h–FFFFFh)
This area is a single 64 KB segment. This segment can be assigned read and write attributes. It is by
default (after reset) read/write disabled and cycles are forwarded to the hub interface. By manipulating
the read/write attributes, the GMCH can “shadow” BIOS into the main DRAM. When disabled, this
segment is not remapped.
4.1.3.
Extended Memory Area
This memory area covers 100000h (1 MB) to FFFFFFFFh (4 GB-1) address range and it is divided into
the following regions:
• Main DRAM Memory from 1 MB to the Top of Memory; maximum of 512 MB using 128M
technology
• PCI Memory space from the Top of Memory to 4 GB with two specific ranges:
• APIC Configuration Space from FEC0_0000h (4 GB-20 MB) to FECF_FFFFh and FEE0_0000h to
FEEF_FFFFh
• High BIOS area from 4 GB to 4 GB–2 MB
Main DRAM Address Range (0010_0000h to Top of Main Memory)
The address range from 1 MB to the top of main memory (TOM) is mapped to main DRAM address
range. The Top of memory is limited to 512 MB. All accesses to addresses within this range will be
forwarded to the DRAM unless a hole in this range is created.
15 MB–16 MB Hole
A hole can be created at 15 MB–16 MB as controlled by the fixed hole enable (FDHC register) in
Device 0 space. Accesses within this hole are forwarded to the hub interface. The range of physical
DRAM memory disabled by opening the hole is not remapped to the Top of the memory – that physical
DRAM space is not accessible. This 15 MB–16 MB hole is an optionally enabled ISA hole. Video
accelerators originally used this hole. It is also used by validation and customer SV teams for some of
their test cards. This is why it is being supported. There is no inherent BIOS request for the 15–16 MB
hole.
Extended SMRAM Address Range (Top of Main Memory–TSEG)
The HSEG and TSEG SMM transaction address spaces reside in this extended memory area.
HSEG
SMM-mode processor accesses to enabled HSEG are remapped to 000A0000h–000BFFFFh. Non-SMMmode processor accesses to enabled HSEG are considered invalid are terminated immediately on the
FSB. The exception to this is non-SMM-mode write-back cycles. They are remapped to SMM space to
maintain cache coherency. AGP and hub interface originated cycles to enabled SMM space are not
allowed. Physical DRAM behind the HSEG transaction address is not remapped and is not accessible.
Datasheet
125
82815 GMCH
R
TSEG
TSEG can be up to 1 MB and is at the top of memory. SMM-mode processor accesses to enabled TSEG
access the physical DRAM at the same address. Non-SMM-mode processor accesses to enabled TSEG
are considered invalid and are terminated immediately on the FSB. The exception is non-SMM-mode
write-back cycles. They are directed to the physical SMM space to maintain cache coherency. AGP and
hub interface originated cycle to enabled SMM space are not allowed.
The size of the SMRAM space is determined by the USMM value in the SMRAM register. When the
extended SMRAM space is enabled, non-SMM processor accesses and all other accesses in this range
are forwarded to the hub interface. When SMM is enabled, the amount of memory available to the
system is equal to the amount of physical DRAM minus the value in the TSEG register.
Note that there is potential for cache corruption if illegal accesses are requested to TSEG. Originally,
TSEG was intended for additional data storage for non-cached solutions. As such, it added protection as
direct reads and writes to TSEG are not allowed to occur outside of SMM. However, when the region is
enabled as cacheable, this protection can cause problems if improperly used. The reason is that, if any
piece of software (including ITP) is to read TSEG outside of SMM, the read can cause corruption of the
cached version of the code in the processor and result in a SMM “hang”.
Example of Problem Manifestation:
1. SMM handler initialized and SMM code/data is written to TSEG
2. Processor cache emptied of TSEG data sometime later as cache lines are evicted and replaced
3. Rogue application requests illegal memory read to TSEG (illegal because processor is not in
SMM)
4. Processor runs memory read cycles to GMCH to perform read from TSEG
5. GMCH realizes processor is NOT in SMM and blocks the reads from targeting actual memory.
Instead it runs the cycle down the hub interface, which ICH2 then converts to a PCI cycle. This
typically gets master aborted and returns a floating bus (FFFFFFFFh).
6. Processor read cycles complete and FFFFFFFFh is pulled into processor cache lines.
7. Processor thinks it has valid TSEG code/data in its cache, when it really has incorrect data
(FFFFFFFFh)
8. Processor runs other system level code, evicting cache lines as needed, but some lines of
FFFFFFFFh remain
9. Eventually, an SMI is generated and this puts the processor into SMM and calls for execution of
the SMM handler stored in TSEG.
10. Processor begins fetching TSEG and hits a line in its cache read by the rogue application
(FFFFFFFFh).
11. This code is corrupted and a “hang” is eminent
The result is that the TSEG protection built into the chipset could potentially cause a system to “hang”
for cached operations, if not properly used. In fact, an application that only reads from the TSEG region
can cause SMRAM corruption by causing the SMM handler to execute bogus code fetched from the PCI
bus.
An alternative is to not use TSEG chipset features at all when running cached. Simply reserve a piece of
system memory at the top of memory region, indicate a lower actual top of memory to the operating
system (through E820h/E801h function calls), and use this region as SMRAM. As there is no restriction
that this memory cannot be accessed when not in SMM mode, then the GMCH will not block accesses to
it. When it is cached, a read to the region (whether performed inside or outside of SMRAM) will return
the correct data and this coherency issue is avoided.
126
Datasheet
82815 GMCH
R
PCI Memory Address Range (Top of Main Memory to 4 GB)
The address range from the top of main DRAM to 4 GB (top of physical memory space supported by the
GMCH) is normally mapped via the hub interface to PCI.
As an internal graphics configuration, there are two exceptions to this rule. Both exception cases are
forwarded to the internal graphics device.
• The first exception is addresses decoded to the local memory range.
• The second exception is addresses decoded to the memory mapped range of the internal graphics
device.
As an AGP configuration, there are two exceptions to this rule.
• Addresses decoded to the AGP memory window defined by the MBASE, MLIMIT, PMBASE, and
PMLIMIT registers are mapped to AGP.
• Addresses decoded to the graphics aperture range defined by the APBASE and APSIZE registers
are mapped to the main DRAM.
There are two sub-ranges within the PCI memory address range defined as APIC Configuration Space
and High BIOS Address Range. As an internal graphics device, the Local Memory Range and the
Memory Mapped Range of the internal Graphics Device MUST NOT overlap with these two ranges.
Similarly, as an AGP device, the AGP memory window and Graphics Aperture Window MUST NOT
overlap with these two ranges. These ranges are described in detail in the following paragraphs.
Datasheet
127
82815 GMCH
R
APIC Configuration Space (FEC0_0000h –FECF_FFFFh,
FEE0_0000h– FEEF_FFFFh)
This range is reserved for APIC configuration space, which includes the default I/O APIC configuration
space. The default Local APIC configuration space is FEE0_0000h to FEEF_0FFFh.
Processor accesses to the local APIC configuration space do not result in external bus activity since the
local APIC configuration space is internal to the processor. However, a MTRR must be programmed to
make the local APIC range uncacheable (UC). The local APIC base address in each processor should be
relocated to the FEC0_0000h (4 GB – 20 MB) to FECF_FFFFh range so that one MTRR can be
programmed to 64 KB for the local and I/O APICs. The I/O APIC(s) usually reside in the I/O Bridge
portion (I/O Controller Hub) of the chipset or as a stand-alone component(s).
I/O APIC units will be located beginning at the default address FEC0_0000h. The first I/O APIC will be
located at FEC0_0000h. Each I/O APIC unit is located at FEC0_x000h where x is I/O APIC unit number
0 through F(hex). This address range will be normally mapped via the hub interface to PCI.
Note:
There is no provision to support an I/O APIC device on AGP
The address range between the APIC configuration space and the High BIOS (FED0_0000h to
FFDF_FFFFh) is always mapped via the hub interface to PCI.
High BIOS Area (FFE0_0000h –FFFF_FFFFh)
The top 2 MB of the extended memory region is reserved for system BIOS (High BIOS), extended BIOS
for PCI devices, and the A20 alias of the system BIOS. The processor begins execution from the High
BIOS after reset. This region is mapped via the hub interface to PCI so that the upper subset of this
region aliases to 16 MB–256 MB range. The actual address space required for the BIOS is less than
2 MB but the minimum processor MTRR range for this region is 2 MB so that full 2 MB must be
considered. The I/O Controller Hub supports a maximum of 1 MB in the High BIOS range.
4.1.3.1.
System Management Mode (SMM) Memory Range
The GMCH supports the use of main memory as System Management RAM (SMRAM) enabling the use
of System Management Mode (SMM). The GMCH supports three SMM options: Compatible SMRAM
(AB segment enabled), High Segment (HSEG), and Top of Memory Segment (TSEG). System
Management RAM (SMRAM) space provides a memory area that is available for the SMI handler’s
code and data storage. This memory resource is normally hidden from the operating system so that the
processor has immediate access to this memory space upon entry to SMM. The GMCH provides three
SMRAM options:
•
Below 1 MB option that supports compatible SMI handlers.
•
Above 1 MB option that allows new SMI handlers to execute with write-back cacheable SMRAM.
•
Optional larger write-back cacheable T_SEG area of either 512 KB or 1MB in size above 1 MB
that is reserved from the highest area in system DRAM memory. The above 1 MB solutions
require changes to compatible SMRAM handler’s code to properly execute above 1 MB.
Refer to the Power Management section for more details on SMRAM support.
Note:
128
The hub interface and AGP masters are not allowed to access the SMM space. This must be insured even
for the GTLB translation.
Datasheet
82815 GMCH
R
4.2.
Memory Shadowing
Any block of memory that can be designated as read only or write-only can be “shadowed” into GMCH
DRAM memory. Typically, this is done to allow ROM code to execute more rapidly out of main DRAM.
ROM is used as a read-only during the copy process while DRAM at the same time is designated writeonly. After copying, the DRAM is designated read-only so that ROM is shadowed. Processor bus
transactions are routed accordingly.
4.3.
I/O Address Space
• The GMCH never responds to I/O cycles initiated on AGP.
• The GMCH does not support the existence of any other I/O devices other than itself on the
processor bus. The GMCH generates either hub interface or AGP/PCI (if enabled) bus cycles for all
processor I/O accesses. If internal graphics is enabled, the GMCH routes the access to hub interface
or legacy I/O registers supported by the internal graphics device. The GMCH contains two internal
registers in the processor I/O space, Configuration Address Register (CONF_ADDR) and the
Configuration Data Register (CONF_DATA). These locations are used to implement PCI
configuration space access mechanism and as described in this document.
• The processor allows 64K+3 bytes to be addressed within the I/O space. The GMCH propagates the
processor I/O address without any translation on to the destination bus and, therefore, provides
addressability for 64K+3 byte locations. Note that the upper 3 locations can be accessed only during
I/O address wrap-around when processor bus A16# address signal is asserted. A16# is asserted on
the processor bus when an I/O access is made to 4 bytes from address 0FFFDh, 0FFFEh, or
0FFFFh. A16# is also asserted when an I/O access is made to 2 bytes from address 0FFFFh.
• The I/O accesses, other than ones used for PCI configuration space access or ones that target the
internal graphics device (or AGP/PCI) are forwarded to the hub interface. The GMCH will not post
I/O write cycles to IDE. The PCICMD1 or PCICMD2 register can disable the routing of I/O cycles
to the AGP.
• The GMCH never responds to I/O cycles initiated on AGP.
4.3.1.
GMCH Decode Rules and Cross-Bridge Address Mapping
The GMCH’s address map applies globally to accesses arriving on any of the three interfaces (i.e., Host
bus, hub interface or from the internal graphics device).
Datasheet
129
82815 GMCH
R
4.3.2.
Address Decode Rules
The GMCH accepts all memory read and write accesses from the hub interface to both system memory
and graphics memory. The hub interface accesses that fall elsewhere within the PCI memory range will
not be accepted. The GMCH never responds to hub interface initiated I/O read or write cycles.
The GMCH accepts accesses from the hub interface to the following address ranges:
• All memory read and write accesses to main DRAM including PAM region (except SMM space)
• All memory read/write accesses to the graphics aperture (DRAM) defined by APBASE and
APSIZE.
• All hub interface memory write accesses to AGP memory range defined by MBASE, MLIMIT,
PMBASE, and PMLIMIT.
• Memory writes to VGA range on AGP, if enabled.
The hub interface memory accesses that fall elsewhere within the memory range are considered invalid
and will be remapped to a translated memory address, snooped on the host bus, and dispatched to
DRAM. Reads will return all 1s with Master Abort completion. Writes will have byte enables deasserted
and will terminate with Master Abort, if completion is required. I/O cycles will not be accepted. They are
terminated with Master Abort completion packets.
The Hub Interface Accesses to GMCH that Cross Device Boundaries
The hub interface accesses are limited to 256 bytes but have no restrictions on crossing address
boundaries. A single hub interface request may, therefore, span device boundaries (AGP, DRAM) or
cross from valid addresses to invalid addresses (or vica versa). The GMCH does not support transactions
that cross device boundaries. For reads and for writes requiring completion, the GMCH provides
separate completion status for each naturally-aligned 32 or 64 byte block. If the starting address of a
transaction hits a valid address, the portion of a request that hits that target device (AGP or DRAM) will
complete normally.
The remaining portion of the access that crosses a device boundary (targets a different device than that of
the starting address) or hits an invalid address will be remapped to memory address 0h, snooped on the
host bus, and dispatched to DRAM. Reads will return all 1s with Master Abort completion. Writes will
have byte enables deasserted and will terminate with Master Abort if completion is required.
If the starting address of a transaction hits an invalid address, the entire transaction will be remapped to
memory address 0h, snooped on the host bus, and dispatched to DRAM. Reads will return all 1s with
Master Abort completion. Writes will have byte enables deasserted and will terminate with Master Abort
if completion is required.
130
Datasheet
82815 GMCH
R
4.3.2.1.
AGP Interface Decode Rules
Cycles Initiated Using PCI Protocol
The GMCH does not support any AGP/PCI access targeting the hub interface. The GMCH will claim
AGP/PCI initiated memory read and write transactions decoded to the main DRAM range or the graphics
Aperture range. All other memory read and write requests will be master-aborted by the AGP/PCI
initiator as a consequence of the GMCH not responding to a transaction.
Under certain conditions, the GMCH restricts access to the DOS compatibility ranges governed by the
PAM registers by distinguishing access type and destination bus. The GMCH accepts AGP/PCI write
transactions to the compatibility ranges if the PAM designates DRAM as write-able. If accesses to a
range are not write enabled by the PAM, the GMCH does not respond and the cycle results in a masterabort. AGP/PCI read transactions to the compatibility ranges are accepted if the PAM designates DRAM
as readable. If accesses to a range are not read enabled by the PAM, the GMCH does not respond and the
cycle will result in a master-abort.
If agent on AGP/PCI issues an I/O or PCI Special Cycle transaction, the GMCH does not respond and
cycle results in a master-abort. The GMCH does not accept PCI configuration cycles to the internal
GMCH devices.
Cycles Initiated Using AGP Protocol
All cycles must reference main memory—main DRAM address range (excluding PAM) or graphics
aperture range (also physically mapped within DRAM but using different address range). AGP accesses
to the PAM region from 640 KB –to- 1 MB are not allowed. AGP accesses to SMM space are not
allowed. AGP-initiated cycles that target DRAM are not snooped on the host bus, even if they fall
outside of the AGP aperture range.
If a cycle is outside of a valid main memory range, then it will terminate as follows:
• Reads: Remap to memory address 0h, return data from address 0h, and set the IAAF error flag.
• Writes: Remapped to memory address 0h with byte enables deasserted (effectively dropped “on the
floor”) and set the IAAF error flag.
AGP Accesses to GMCH that CrosbvDevice Boundaries
For FRAME# accesses, when an AGP or PCI master gets disconnected, it will resume at the new address
that allows the cycle to be routed to or claimed by the new target. Therefore, accesses should be
disconnected by the target on potential device boundaries. The GMCH disconnects AGP/PCI
transactions on 4 KB boundaries.
AGP PIPE# and SBA accesses are limited to 256 bytes and must hit DRAM. AGP accesses are
dispatched to DRAM on naturally aligned 32-byte block boundaries. The portion of the request that hits
a valid address completes normally. The portion of a read access that hits an invalid address is remapped
to address 0h, returns data from address 0h, and sets the IAAF error flag. The portion of a write access
that hits an invalid address is remapped to memory address 0h with byte enables deasserted (effectively
dropped “on the floor”) and set the IAAF error flag.
Datasheet
131
82815 GMCH
R
4.3.2.2.
Legacy VGA Ranges
The legacy VGA memory range A0000h–BFFFFh is mapped either to the hub interface or to AGP/PCI1
depending on the programming of the VGA Enable bit in the BCTRL configuration register in GMCH
Device #1 configuration space, and the MDAP bit in the GMCHCFG configuration register in Device #0
configuration space. The same register controls mapping VGA I/O address ranges. The VGA I/O range is
defined as addresses where A[9:0] are in the ranges 3B0h to 3BBh and 3C0h to 3DFh (inclusive of ISA
address aliases: A[15:10] are not decoded). The function and interaction of these two bits is described
below:
MDA Present (MDAP): This bit works with the VGA Enable bit in the BCTRL register of device 1 to
control the routing of processor-initiated transactions targeting MDA compatible I/O and memory
address ranges. This bit should not be set when the VGA Enable bit is not set. If the VGA enable bit is
set, accesses to IO address range x3BCh–x3BFh are forwarded to the hub interface. If the VGA enable
bit is not set, I/O address range accesses x3BCh–x3BFh are treated like other I/O accesses (the cycles are
forwarded to AGP if the address is within IOBASE and IOLIMIT and ISA enable bit is not set);
otherwise, they are forwarded to the hub interface. MDA resources are defined as the following:
Memory:
0B0000h–0B7FFFh
I/O:
3B4h, 3B5h, 3B8h, 3B9h, 3BAh, 3BFh,
(including ISA address aliases, A[15:10] are not used in decode)
Any I/O reference that includes the I/O locations listed above, or their aliases, are forwarded to the hub
interface, even if the reference includes I/O locations not listed above.
VGA Enable: Controls the routing of processor-initiated transactions targeting VGA compatible I/O and
memory address ranges. When this bit is set, the GMCH forwards the following processor accesses to
AGP:
• Memory accesses in the range 0A0000h to 0BFFFFh
• I/O addresses where A[9:0] are in the ranges 3B0h to 3BBh and 3C0h to 3DFh
(inclusive of ISA address aliases: A[15:10] are not decoded)
When this bit is set , forwarding of these accesses issued by the processor is independent of the I/O
address and memory address ranges defined by the previously defined Base and Limit registers.
Forwarding of these accesses is also independent of the settings of bit 2 (ISA Enable) of BCTRL if this
bit is 1. If the VGA enable bit is set, accesses to I/O address range x3BCh–x3BFh are forwarded to the
hub interface. If the VGA enable bit is not set, I/O address range accesses x3BCh–x3BFh are treated like
other I/O accesses (the cycles are forwarded to AGP, if the address is within IOBASE and IOLIMIT and
ISA enable bit is not set); otherwise, they are forwarded to the hub interface.
If this bit is 0 (default), VGA compatible memory and I/O range accesses are not forwarded to AGP;
rather, they are mapped to the hub interface, unless they are mapped to AGP via I/O and memory range
registers defined above (IOBASE, IOLIMIT, MBASE, MLIMIT, PMBASE, PMLIMIT).
The following table shows the behavior for all combinations of MDA and VGA:
132
VGA
MDA
0
0
Behavior
All references to MDA and VGA go to the hub interface
0
1
Illegal combination (DO NOT USE)
1
0
All references to VGA Go To AGP/PCI. MDA-only references (I/O Address 3BFh and
aliases) will go to the hub interface.
1
1
VGA references go to AGP/PCI; MDA references go to the hub interface.
Datasheet
82815 GMCH
R
4.4.
Host Interface
The host interface of the GMCH is optimized to support the Intel Pentium III processor,
Intel Pentium II processor, and Intel CeleronTM processor. The GMCH implements the host address,
control, and data bus interfaces within a single device. The GMCH supports a 4-deep in-order queue
(i.e., supports pipelining of up to 4 outstanding transaction requests on the host bus) . Host bus addresses
are decoded by the GMCH for accesses to system memory, PCI memory ane PCI I/O (via hub interface),
PCI configuration space, and graphics memory. The GMCH takes advantage of the pipelined addressing
capability of the processor to improve the overall system performance. The GMCH supports the 370-pin
socket processor connector.
4.4.1.
Host Bus Device Support
The GMCH recognizes and supports a large subset of the transaction types that are defined for the Intel
Pentium III processor, Intel Pentium II processor, or Intel CeleronTM processor bus interface. However,
each of these transaction types have a multitude of response types, some of which are not supported by
this controller. All transactions are processed in the order that they are received on the processor bus.
Table 10. Summay of Transactions Supported By GMCH
Transaction
Datasheet
REQa[4:0]#
REQb[4:0]#
GMCH Support
Deferred Reply
00000
XXXXX
The GMCH will initiate a deferred reply request for a
previously deferred transaction.
Reserved
00001
XXXXX
Reserved
Interrupt
Acknowledge
01000
00000
Interrupt acknowledge cycles are forwarded to the hub
interface.
Special
Transactions
01000
00001
See separate table in special cycles section.
Reserved
01000
0001x
Reserved
Reserved
01000
001xx
Reserved
Branch Trace
Message
01001
00000
The GMCH will terminate a branch trace message without
latching data.
Reserved
01001
00001
Reserved
Reserved
01001
0001x
Reserved
Reserved
01001
001xx
Reserved
I/O Read
10000
0 0 x LEN#
I/O read cycles are forwarded to hub interface. I/O cycles
that are in the GMCH configuration space are not forwarded
to the hub interface.
I/O Write
10001
0 0 x LEN#
I/O write cycles are forwarded to hub interface. I/O cycles
that are in the GMCH configuration space are not forwarded
to the hub interface.
Reserved
1100x
00xxx
Reserved
Memory Read &
Invalidate
00010
0 0 x LEN#
Host initiated memory read cycles are forwarded to DRAM
or the hub interface.
Reserved
00011
0 0 x LEN#
Reserved
133
82815 GMCH
R
Transaction
REQa[4:0]#
REQb[4:0]#
GMCH Support
Memory Code
Read
00100
0 0 x LEN#
Memory code read cycles are forwarded to DRAM or the
hub interface.
Memory Data
Read
00110
0 0 x LEN#
Host-initiated memory read cycles are forwarded to DRAM
or the hub interface.
Memory Write
(no retry)
00101
0 0 x LEN#
This memory write is a writeback cycle and cannot be
retried. The GMCH forwards the write to DRAM.
Memory Write
(can be retried)
00111
0 0 x LEN#
The standard memory write cycle is forwarded to DRAM or
the hub interface.
NOTES:
1. For Memory cycles, REQa[4:3]# = ASZ#. GMCH only supports ASZ# = 00 (32 bit address).
2. REQb[4:3]# = DSZ#. DSZ# = 00 (64 bit data bus size).
3. LEN# = data transfer length as follows:
LEN#
Data length
00
<= 8 bytes (BE[7:0]# specify granularity)
01
Length = 16 bytes BE[7:0]# all active
10
Length = 32 bytes BE[7:0]# all active
11
Reserved
Table 11. Host Responses Supported by the GMCH
134
RS2#
RS1#
RS0#
Description
GMCH Support
0
0
0
idle
0
0
1
Retry Response
This response is generated if an access is to a resource
that cannot be accessed by the processor at this time and
the logic must avoid deadlock . Hub interface directed
reads, writes, and DRAM locked reads can be retried.
0
1
0
Deferred
Response
This response can be returned for all transactions that can
be executed ‘out of order.’ Hub interface directed reads
(memory, I/O and Interrupt Acknowledge) and writes (I/O
only), and internal Graphics device directed reads
(memory and I/O) and writes (I/O only) can be deferred.
0
1
1
Reserved
Reserved
1
0
0
Hard Failure
Not supported.
1
0
1
No Data
Response
This is for transactions where the data has already been
transferred or for transactions where no data is
transferred. Writes and zero length reads receive this
response.
1
1
0
Implicit
Writeback
This response is given for those transactions where the
initial transactions snoop hits on a modified cache line.
1
1
1
Normal Data
Response
This response is for transactions where data accompanies
the response phase. Reads receive this response.
Datasheet
82815 GMCH
R
4.4.2.
Special Cycles
A Special Cycle is defined when REQa[4:0] = 01000 and REQb[4:0]= xx001. The first address phase
Aa[35:3]# is undefined and can be driven to any value. The second address phase, Ab[15:8]# defines the
type of Special Cycle issued by the processor.
Table 12 specifies the cycle type and definition as well as the action taken by the GMCH when the
corresponding cycles are identified.
Table 12. Special Cycles
BE[7:0]#
Datasheet
Special Cycle Type
Action Taken
0000 0000
NOP
This transaction has no side-effects.
0000 0001
Shutdown
This transaction is issued when an agent detects a severe software error
that prevents further processing. This cycle is claimed by the GMCH. The
GMCH issues a shutdown special cycle on the hub interface. This cycle is
retired on the processor bus after it is terminated on the hub interface via
a master abort mechanism.
0000 0010
Flush
This transaction is issued when an agent has invalidated its internal
caches without writing back any modified lines. The GMCH claims this
cycle and retires it.
0000 0011
Halt
This transaction is issued when an agent executes a HLT instruction and
stops program execution. This cycle is claimed by the GMCH and
propagated to the hub interface as a Special Halt Cycle. This cycle is
retired on the processor bus after it is terminated on the hub interface via
a master abort mechanism.
0000 0100
Sync
This transaction is issued when an agent has written back all modified
lines and has invalidated its internal caches. The GMCH claims this cycle
and retires it.
0000 0101
Flush Acknowledge
This transaction is issued when an agent has completed a cache sync
and flush operation in response to an earlier FLUSH# signal assertion.
The GMCH claims this cycle and retires it.
0000 0110
Stop Clock
Acknowledge
This transaction is issued when an agent enters Stop Clock mode. This
cycle is claimed by the GMCH and propagated to the hub interface as a
Special Stop Grant Cycle. This cycle is completed on the processor bus
after it is terminated on the hub interface via a master abort mechanism.
0000 0111
SMI Acknowledge
This transaction is first issued when an agent enters the System
Management Mode (SMM). Ab[7]# is also set at this entry point. All
subsequent transactions from the processor with Ab[7]# set are treated
by the GMCH as accesses to the SMM space. No corresponding cycle is
propagated to the hub interface. To exit the System Management Mode
the processor issues another one of these cycles with the Ab[7]# bit
deasserted. The SMM space access is closed by the GMCH at this point.
All others
Reserved
135
82815 GMCH
R
4.5.
System Memory DRAM Interface
The GMCH integrates a system DRAM controller that supports a 64-bit DRAM array. The DRAM type
supported is synchronous (SDRAM). The GMCH generates the SCSA#, SCSB#, SDQM, SCAS#,
SRAS#, SWE# and multiplexed addresses, SMA for the DRAM array. The GMCH’s DRAM interface
operates at a clock frequency of either 100 or 133 MHz, dependent upon the system bus interface clock
frequency. The DRAM controller interface is fully configurable through a set of control registers.
Complete descriptions of these registers are given in the register description section of this document.
The GMCH supports industry standard 64-bit wide DIMM modules with SDRAM devices. The 2 bank
select lines (SBS[1:0]), the 12 address lines (SMAA[11:0]), and copies of 4 address lines (SMAB[7:4]#
and SMAC[7:4]#) allow the GMCH to support 64-bit wide DIMMs using 16Mb, 64Mb, or 128Mb
technology SDRAMs. The GMCH has a sufficient amount of SCS# lines to enable the support of up to
six 64-bit rows of DRAM. For write operations of less than a QWord, the GMCH performs a byte-wise
write. The GMCH targets SDRAM with CL2 and CL3 and supports both single and double-sided
DIMMs. The GMCH provides refresh functionality with programmable rate (normal DRAM rate is
1 refresh/15.6 µs). The GMCH can be configured via the Page Closing Policy Bit in the GMCH
Configuration Register to keep multiple pages open within the memory array. Pages can be kept open in
any one row of memory. Up to 4 pages can be kept open within that row (The GMCH only supports 4
Bank SDRAMs on system DRAM interface).
4.5.1.
DRAM Organization and Configuration
The GMCH supports 64-bit DRAM configurations. In the following discussion the term row refers to a
set of memory devices that are simultaneously selected by a SCS# signal. The GMCH supports a
maximum of 6 rows of memory. Both single-sided and double-sided DIMMs are supported.
The interface consists of the following pins:
Multiple copies:
SMAA[7:4], SMAB[7:4]# , SMAC[7:4]#
Single Copies:
SMD[63:0]
SDQM[7:0]
SMAA[12:8,3:0]
SBS[1:0]
SCSA[5:0]#
SCSB[5:0]#
SCAS#
SRAS#
SWE#
SCKE[1:0]
The GMCH supports DIMMs populated with 8, 16, and 32 bit wide SDRAM devices. Registered
DIMMs or DIMMs populated with 4 bit wide SDRAM devices are not supported. The GMCH supports
3.3V standard SDRAMs.
136
Datasheet
82815 GMCH
R
Table 13 illustrates a sample of the possible DIMM socket configurations along with corresponding DRP
programming. See the register section of this document for a complete DRP programming table.
Table 13. Sample Of Possible Mix And Match Options For 4 Row/2 DIMM Configurations
DIMM0
DIMM1
DRP
Total Memory
0
4x(4M x 16 ) S
70
32 MB
4x (4M x16 ) S
0
07
32 MB
4x(4Mx16) + 2x(2Mx32) D
0
08
48 MB
4x(4Mx16) S
4x(4Mx16) S
77
64 MB
8x(8Mx8) + 4x(4Mx16) D
0
0B
96 MB
8x(8Mx8) D
0
0C
128 MB
8x(8Mx8) D
8x(8Mx8) D
CC
256 MB
NOTES:
1. “S” denotes single-sided DIMMs, “D” denotes double-sided DIMMs.
4.5.1.1.
Configuration Mechanism For DIMMs
Detection of the type of DRAM installed on the DIMM is supported via Serial Presence Detect
mechanism as defined in the JEDEC 168-pin DIMM standard. This standard uses the SCL, SDA and
SA[2:0] pins on the DIMMs to detect the type and size of the installed DIMMs. No special
programmable modes are provided on the GMCH for detecting the size and type of memory installed.
Type and size detection must be done via the serial presence detection pins. Use of Serial Presence
Detection is required.
Memory Detection and Initialization
Before any cycles to the memory interface can be supported, the GMCH DRAM registers must be
initialized. The GMCH must be configured for operation with the installed memory types. Detection of
memory type and size is done via the System Management Bus (SMBus) interface on the I/O Controller
Hub. This two wire bus is used to extract the DRAM type and size information from the serial presence
detect port on the DRAM DIMM modules.
DRAM DIMM modules contain a 5-pin serial presence detect interface, including SCL (serial clock),
SDA (serial data) and SA[2:0]. Devices on the SMBus bus have a seven bit address. For the DRAM
DIMM modules, the upper four bits are fixed at 1010. The lower three bits are strapped on the SA[2:0]
pins. SCL and SDA are connected directly to the System Management Bus on the I/O Controller Hub.
Thus, data is read from the Serial Presence Detect port on the DRAM DIMM modules via a series of IO
cycles to the I/O Controller Hub. BIOS essentially needs to determine the size and type of memory used
for each of the four rows of memory in order to properly configure the GMCH system memory interface.
SMBus Configuration and Access of the Serial Presence Detect Ports
For more details on this, see the Intel® 82801AA (ICH) and Intel® 82801AB (ICH0) I/O Controller Hub
datasheet or Intel® 82801BA (ICH2) I/O Controller Hub datasheet..
Datasheet
137
82815 GMCH
R
4.5.1.2.
DRAM Register Programming
This section provides an overview of how the required information for programming the DRAM registers
is obtained from the Serial Presence Detect ports on the DIMMs. The Serial Presence Detect ports are
used to determine Refresh Rate, MA and MD Buffer Strength, Row Type (on a row by row basis),
SDRAM Timings, Row Sizes and Row Page Sizes. Table 14 lists a subset of the data available through
the on-board Serial Presence Detect ROM on each DIMM module.
Table 14. Data Bytes on DIMM Used for Programming DRAM Registers
Byte
Function
2
Memory Type (EDO, SDRAM) the GMCH only supports SDRAM.
3
# of Row Addresses, not counting Bank Addresses
4
# of Column Addresses
5
# of banks of DRAM (Single or Double sided) DIMM
12
Refresh Rate
17
# Banks on each SDRAM Device
36–41
42
Access Time from Clock for CAS# Latency 1 through 7
Data Width of SDRAM Components
Table 14 is only a subset of the defined SPD bytes on the DIMM module. These bytes collectively
provide enough data for BIOS to program the GMCH DRAM registers.
4.5.2.
DRAM Address Translation and Decoding
The GMCH translates the address received on the host bus, hub interface, or from the internal graphics
device to an effective memory address. The GMCH supports 16 Mbit and 64 Mbit SDRAM devices. The
GMCH supports a 2 KB page sizes only. The multiplexed row / column address to the DRAM memory
array is provided by the SBS[1:0] and SMAA[11:0] signals and copies. These addresses are derived from
the host address bus as defined by by the following table for SDRAM devices.
• Row size is internally computed using the values programmed in the DRP register.
• Up to 4 pages can be open at any time within any row (Only 2 active pages are supported in rows
populated with either 8 MBs or 16 MBs ).
138
Datasheet
82815 GMCH
R
16
64
64
128
128
256
256
2M
8M
4M
16M
8M
32M
16M
NOTES:
4.5.3.
8
8
16
8
16
8
16
Row
Col
Bank
11
9
1
12
12
12
12
13
13
9
2
8
2
10
2
9
2
10
2
9
2
16
64
32
128
64
256
128
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
Mem
Size
(MB)
BS
Address Usage
BS
Width
Tech
(Mb)
Depth
Table 15. GMCH DRAM Address Mux Function
1
0
12
11
10
9
8
7
6
5
4
3
2
1
0
X
11
X
X
[A]
22
21
20
19
18
17
16
15
14
13
X
11
X
X
PA
X
23
10
9
8
7
6
5
4
3
12
11
X
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
X
25
10
9
8
7
6
5
4
3
12
11
X
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
X
X
10
9
8
7
6
5
4
3
12
11
X
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
26
25
10
9
8
7
6
5
4
3
12
11
X
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
X
25
10
9
8
7
6
5
4
3
12
11
27
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
26
25
10
9
8
7
6
5
4
3
12
11
26
24
[A]
22
21
20
19
18
17
16
15
14
13
12
11
X
X
PA
X
25
10
9
8
7
6
5
4
3
MA bit 10 at RAS time uses the XOR of Address bit 12 and Address bit 23
DRAM Array Connectivity
Figure 6. DRAM Array Sockets
SCS[3:2]#
SCS[1:0]#
SCKE0
Note:
Min (16Mbit) 8MB
Max (64Mbit) 256MB
Max (128Mbit) 512MB
SCKE1
SRAS#
SCAS#
SWE#
SBS[1:0]
SMAA[12:8,3:0]
SMAA[7:4]
SMAB[7:4]#
SDQM[7:0]
SMD[63:0]
DIMM_CLK[3:0]
DIMM_CLK[7:4]
SMB_CLK
SMB_DATA
mem_dimm
Datasheet
139
82815 GMCH
R
4.5.4.
SDRAMT Register Programming
Several DRAM timing parameters are programmable in the GMCH configuration registers. Table 16
summarizes the programmable parameters.
Table 16. Programmable SDRAM Timing Parameters
Parameter
DRAMT Bit
Values (SCLKs)
RAS# Precharge (SRP)
0
2,3
RAS# to CAS# Delay (SRCD)
1
2,3
CAS# Latency (CL)
2
2,3
DRAM Cycle Time (DCT)
4
Tras = 5,6
Trc = 7,8
These parameters are controlled via the DRAMT register. To support different device speed grades,
CAS# Latency, RAS# to CAS# Delay, and RAS# Precharge are all programmable as either two or three
SCLKs. To provide flexibility, these are each controlled by separate register bits (i.e., the GMCH can
support any combination of CAS# Latency, RAS#-to-CAS# Delay and RAS# Precharge).
4.5.5.
SDRAM Paging Policy
The GMCH can maintain up to 4 active pages in any one row; however, the GMCH does not support
active pages in more than 1 row at a time.
The DRAM page closing policy (DPCP) in the GMCH configuration register (GMCHCFG) controls the
page closing policy of the GMCH. This bit controls whether the GMCH “precharges bank” or
“precharges all” during the service of a page miss. When this bit is 0, the GMCH prechanges bank during
the service of a page miss. When this bit is 1, the GMCH prechanges all during the service of a page
miss.
4.6.
Intel Dynamic Video Memory Technology (D.V.M.T.)
The internal graphics device on the GMCH supports Intel Dynamic Video Memory Technology
(D.V.M.T.). D.V.M.T. dynamically responds to application requirements by allocating the proper
amount of display and texturing memory. For more details, refer to the document entitled, “Intel 810
Chipset: Great Performance for Value PCs” available at:
http://developer.intel.com/design/chipsets/810/810white.htm.
In addition to D.V.M.T., the GMCH supports Display Cache (DC). The graphics engine of the GMCH
uses DC for implementing rendering buffers (e.g., Z buffers). This rendering model requires 4 MB of
display cache and allows graphics rendering (performed across the graphics display cache bus) and
texture MIP map access (performed across the system memory bus) simultaneously. In using D.V.M.T.,
all graphics rendering is implemented in system memory. The system memory bus is arbitrated between
texture MIP-map accesses and rendering functions.
140
Datasheet
82815 GMCH
R
4.7.
Display Cache Interface
The GMCH Display Cache (DC) is a single channel 32-bit wide SDRAM interface. The GMCH handles
the control and timing for the display cache. The display cache interface of the GMCH generates the
LCS#, LDQM[7:0], LSCAS#, LSRAS#, LWE#, LMD[31:0] and multiplexed addresses, LMA[11:0] for
the display cache DRAM array. The GMCH also generates the clock LTCLK[1:0] for write cycles as
well as LOCLK for read cycle timings.
The display cache interface of the GMCH supports single data rate synchronous dynamic random access
memory (SDRAM). It supports a single 32-bit wide memory channel. The interface handles the operation
of D.V.M. with DC at 133 MHz. The DRAM controller interface is fully configurable through a set of
control registers.
Internal buffering (FIFOs) of the data to and from the display cache ensures the synchronization of the
data to the internal pipelines. The D.V.M. with DC interface clocking is divided synchronous with
respect to the core and system bus.
The display cache resides on an AGP In-Line Memory Module (AIMM). The startup sequencing for the
AIMM (which is interfaced via the AGP connector), is as follows:
1. System BIOS detects if an AGP card is present by performing a configuration read to PCI. If an
AGP card is present, it becomes the display device and bit 0 of the APCONT register should be set
to 0. No further initialization of internal graphics will take place. If internal graphics is the
preferred display device, bit 0 of the APCONT register should be set to 1. If no AGP card is
present, the internal graphics becomes the display device and bit 0 of the APCONT register should
be set to 1. PCI enumeration takes place at this point.
In the case where internal graphics is selected, the remaining steps still apply:
2.
System BIOS determines if an AIMM card (local memory) is present.
If the AIMM card is present, the following steps take place:
3.
4.
4.7.1.
Local Memory Clock Frequency is determined with a reset strap (on AGP pin SBA[7]) sampled as
an input during reset.
Memory Timing Options will be determined empirically by the system BIOS. BIOS will start with
programming slow timings (CAS Latency, RAS Pre-charge, etc.) and then trying faster timings
until it breaks. The settings that optimize performance without compromising functionality will be
selected.
Supported DRAM Types for Display Cache Memory
The GMCH supports 1Mx16 and 2Mx32 SDRAMs; however, the GMCH only supports 4 MB of display
cache.
Datasheet
141
82815 GMCH
R
4.7.2.
Memory Configurations
Table 17 provides a summary of the characteristics of memory configurations supported. The GMCH
supports a 32-bit wide channel populated with a single row of 1Mx16 SDRAMs.
Table 17. Memory Size for each configuration :
SDRAM
SDRAM
SDRAM
# of
Address Size
DRAM
Tech.
Density
Width
Banks
Bank
Row
Column
Addressing
16 Mbit
1M
16
2
1
11
8
Asymmetric
16 Mbit
1M
32
DRAM Size
4 MB
The following figure shows the GMCH LMI connected to 4 MB of memory in a 32-bit SDRAM channel
configuration.
Figure 7. GMCH Display Cache Interface to 4 MB
GMCH
LCS#
LSRAS
LSCAS
LWE#
LMA[11:0]
LMD[31:0]
LDQM[3:0]
LTCLK[1:0]
LRCLK
LOCLK
(2) 1M x 16
SDRAM
disp_ca_interface
142
Datasheet
82815 GMCH
R
4.7.3.
Address Translation
The GMCH contains address decoders that translate the address received by the display cache into an
effective display cache address. The LMA[11:0] bits are as defined below. Entries in the table
(e.g., A21(X)) imply that the GMCH puts out A21 on that MA line but it is not used by the SDRAM.
Table 18. GMCH Local Memory Address Mapping
MA
1Mx16
Row
Column
11(BA)
A10
A10
10
A11
X
9
A21
X
8
A20
X
7
A19
A9
6
A18
A8
5
A17
A7
4
A16
A6
3
A15
A5
2
A14
A4
1
A13
A3
0
A12
A2
BA = Bank address
4.7.4.
Display Cache Interface Timing
The GMCH provides a variety of programmable wait states for DRAM read and write cycles. These
options are programmed in the display cache I/O addresses of the GMCH configuration space. The wrap
type and the burst length is implied since they are not programmable and fixed. Only sequential wrap is
allowed. Burst length is fixed at two.
Datasheet
143
82815 GMCH
R
4.8.
Internal Graphics Device
4.8.1.
3D/2D Instruction Processing
The GMCH contains an extensive set of instructions that control various functions including 3D
rendering, BLT and STRBLT operations, display, motion compensation, and overlay. The 3D
instructions set 3D pipeline states and control the processing functions. The 2D instructions provide an
efficient method for invoking BLT and STRBLT operations.
The graphics controller executes instructions from one of two instruction buffers located in system
memory: Interrupt Ring or Low Priority Ring. Instead of writing instructions directly to the GMCH’s
graphics controller, software sets up instruction packets in these memory buffers and then instructs the
GMCH to process the buffers. The GMCH uses DMAs to put the instructions into its FIFO and executes
them. Instruction flow in the ring buffer instruction stream can make calls to other buffers, much like a
software program makes subroutine calls. Flexibility has been built into the ring operation permitting
software to efficiently maintain a steady flow of instructions.
Batching instructions in memory ahead of time and then instructing the graphics controller to process the
instructions provides significant performance advantages over writing directly to FIFOs including: 1)
Reduced software overhead, 2) Efficient DMA instruction fetches from graphics memory, and 3)
Software can more efficiently set up instruction packets in buffers in graphics memory (faster than
writing to FIFOs).
Figure 8. 3D/2D Pipeline Preprocessor
Low Priority Ring
(Graphics Memory)
Batch Buffers
Instruction
Batch Buff Instr
Instruction
DMA
Instruction access and decoding
2D Instructions
DMA
FIFO
Interrupt Ring
(Graphics Memory)
Batch Buffers
Instr
Parser
3D Instructions (3D state,
3D Primitives, STRBLT,
Motion Compensation)
Instruction
DMA
BLT
Engine
3D
Engine
Batch Buff Instr
Instruction
Display
Engine
Overlay
Engine
cmd_str.vsd
144
Datasheet
82815 GMCH
R
4.8.2.
3D Engine
The 3D engine of the GMCH has been architected as a deep pipeline, where performance is maximized
by allowing each stage of the pipeline to simultaneously operate on different primitives or portions of the
same primitive. The GMCH supports perspective-correct texture mapping, bilinear, trilinear, and
anisotropic MIP map filtering, Gouraud shading, alpha-blending, colorkeying and chromakeying, full
color specular shading, fogging and Z Buffering. These features can be independently enabled or
disabled via a set of 3D instructions. In addition, the GMCH supports a Dynamic Video Memory
Technology (D.V.M.T.) that allows the entire 3D rendering process to take place in system memory;
thus, alleviating the need for the display cache.
The main blocks of the pipeline are the Setup Engine, Scan Converter, Texture Pipeline, and Color
Calculator block. A typical programming sequence would be to send instructions to set the state of the
pipeline followed by rendering instructions containing 3D primitive vertex data.
4.8.3.
Buffers
The 2D, 3D, and video capabilities of the GMCH provide control over a variety of graphics buffers that
can be implemented either in display cache or system memory. To aid the rendering process, the display
cache of the GMCH contains two hardware buffers—the Front Buffer (display buffer) and the Back
Buffer (rendering buffer). The image being drawn is not visible until the scene is complete and the back
buffer made visible (via an instruction) or copied to the front buffer (via a 2D BLT operation). By
rendering to one and displaying from the other, the possibility of image tearing is removed. This also
speeds up the display process over a single buffer.
The 3D pipeline of the GMCH operates on the Back Buffer and the Z Buffer. The pixels’ 16-bit
(or 15-bit) RGB colors are stored in the back buffer. The Z-buffer can be used to store 16-bit depth
values or 5-bit “destination alpha” values. The instruction set of the GMCH provides a variety of controls
for the buffers (e.g., initializing, flip, clear, etc.).
Figure 9. Data Flow for the 3D Pipeline
GMCH Graphics Pipeline
(Conceptual Representation)
System Memory
Instructions
and Data
Textures
Setup
Frame Buffer
Discard
(Back Face
Culling)
Primitives
Pixels
Mapping
Engine
Display Cache
Color
Calculator
Rasterize
GMCH
Interface
Depth Buffer
(Z-Buffer)
Notes:
1. Frame Buffer = Front and Back Buffers
Datasheet
3d_pipe2.vsd
145
82815 GMCH
R
4.8.4.
Setup
The setup stage of the pipeline takes the input data associated with each vertex of the line or triangle
primitive and computes the various parameters required for scan conversion. In formatting this data, the
GMCH maintains sub-pixel accuracy. Data is dynamically formatted for each rendered polygon and
output to the proper processing unit. As part of the setup, the GMCH removes polygons from further
processing, if they are not facing the user’s viewpoint (referred to as “ Back Face Culling”).
4.8.5.
Texturing
The GMCH allows an image, pattern, or video to be placed on the surface of a 3D polygon. Textures
must be located in system memory. Being able to use textures directly from system memory means that
large complex textures can easily be handled without the limitations imposed by the traditional approach
of only using the display cache.
The texture processor receives the texture coordinate information from the setup engine and the texture
blend information from the scan converter. The texture processor performs texture color or chroma-key
matching, texture filtering (anisotropic, bilinear, and trilinear interpolation), and YUV to RGB
conversions.
The GMCH supports up to 11 Levels-of-Detail (LODs) ranging in size from 1024x1024 to 1x1 texels.
(A texel is defined as a texture map pixel). Textures need not be square. Included in the texture processor
is a small cache that provides efficient mip-mapping.
• Nearest. Texel with coordinates nearest to the desired pixel is used. (This is used if only one LOD
is present).
• Linear. A weighted average of a 2x2 area of texels surrounding the desired pixel are used. (This is
used if only one LOD is present).
• Mip Nearest. This is used if many LODs are present. The appropriate LOD is chosen and the texel
with coordinates nearest to the desired pixel are used.
• Mip Linear. This is used if many LODs are present. The appropriate LOD is chosen and a weighted
average of a 2x2 area of texels surrounding the desired pixel are used. This is also referred to as bilinear mip-mapping.
• Trilinear. Tri-linear filtering blends two mip maps of the same image to provide a smooth transition
between different mips (floor and ceiling of the calculated LOD).
• Anisotropic. This can be used if multiple LODs are present. This filtering method improves the
visual quality of texture-mapped objects when viewed at oblique angles (i.e., with a high degree of
perspective foreshortening). The improvement comes from a more accurate (anisotropic) mapping
of screen pixels onto texels – where using bilinear or trilinear filtering can yield overly-blurred
results. Situations where anisotropic filtering demonstrates superior quality include text viewed at an
angle, lines on roadways, etc.
146
Datasheet
82815 GMCH
R
The GMCH can store each of the above mip-maps in any of the following formats:
• 8bpt Surface Format
• 16bpt Surface Format
 RGB 565
 ARGB 1555
 ARGB 4444
 AY 88
• 8bpt (Indexed) Surface Format
 RGB 565
 ARGB 1555
 ARGB 4444
 AY 88
• 4:2:2
 YcrCb, Swap Y Format
 YcrCb, Normal
 YcrCb, UV Swap
 YcrCb, UV/Y Swap
Many texture mapping modes are supported. Perspective correct mapping is always performed. As the
map is fitted across the polygon, the map can be tiled, mirrored in either the U or V directions, or
mapped up to the end of the texture and no longer placed on the object (this is known as clamp mode).
The way a texture is combined with other object attributes is also definable.
Texture ColorKey and ChromaKey
ColorKey and ChromaKey describe two methods of removing a specific color or range of colors from a
texture map before it is applied to an object. For “nearest” texture filter modes, removing a color simply
makes those portions of the object transparent (the previous contents of the back buffer show through).
For “ linear “ texture filtering modes, the texture filter is modified if only the non-nearest neighbor texels
match the key (range).
ColorKeying occurs with paletted textures, and removes colors according to an index (before the palette
is accessed). When a color palette is used with indices to indicate a color in the palette, the indices can be
compared against a state variable “ColorKey Index Value” and if a match occurs and ColorKey is
enabled, then this value’s contribution is removed from the resulting pixel color. The GMCH defines
index matching as ColorKey.
ChromaKeying can be performed for both paletted and non-paletted textures, and removes texels that fall
within a specified color range. The ChromaKey mode refers to testing the RGB or YUV components to
see if they fall between high and low state variable values. If the color of a texel contribution is in this
range and chromaKey is enabled, then this contribution is removed from the resulting pixel color.
Multiple Texture Composition
The GMCH includes support for two simultaneous texture maps. This support greatly reduces the need
for multipass compositing techniques for effects such as diffuse light maps, specular reflection maps,
bump mapping, detail textures, gloss maps, shadows, and composited effects like dirt or tire marks.
Supporting these techniques in hardware greatly increases compositing performance by reducing the need
to read and write the frame buffer multiple times.
Datasheet
147
82815 GMCH
R
This multitexture support provides a superset of the “legacy” one-texture (pre-DirectX 6) texture blend
modes and a large subset of the operations defined in DirectX 6 and the OpenGL ARB multitexture
extensions.
The Multitexture Compositing Unit is capable of combining the interpolated vertex diffuse color, a
constant color value, and up to two texels per pixel in a fully-programmable fashion. Up to three
operations (combinations) can be performed in a pipelined organization, with intermediate storage to
support complex equations (e.g., of the form “A*B + C*D) required for light maps and specular gloss
maps. Separate operations can be performed on color (RGB) and alpha components.
4.8.6.
2D Operation
The GMCH contains BLT and STRBLT functionality, a hardware cursor, and an extensive set of 2D
registers and instructions.
GMCH VGA Registers and Enhancements
The 2D registers are a combination of registers defined by IBM* when the Video Graphics Array (VGA)
was first introduced, and others that Intel has added to support graphics modes that have color depths,
resolutions, and hardware acceleration features that go beyond the original VGA standard. The GMCH
improves upon VGA by providing additional features that are used through numerous additional
registers.
The GMCH also supports an optional display cache. As an improvement on the VGA standard display
cache port-hole, the GMCH also maps the entire display cache into part of a single contiguous memory
space at a programmable location, providing what is called “linear” access to the display cache. The size
of this memory can be up to 4 MB, and the base address is set via PCI configuration registers.
Alternatively, these buffers may be implemented in system memory (via D.V.M.), thus alleviating the
need for the display cache.
4.8.7.
Fixed Blitter (BLT) and Stretch Blitter (STRBLT) Engines
The GMCH‘s 64-bit BLT engine provides hardware acceleration for many common Windows*
operations. The following are two primary BLT functions: Fixed Blitter (BLT) and Stretch Blitter
(STRBLT). The term BLT refers to a block transfer of pixel data between memory locations. The word
“fixed” is used to differentiate from the Stretch BLT engine. The BLT engine can be used for the
following:
• Move rectangular blocks of data between memory locations
• Data alignment
• Perform logical operations
The GMCH has instructions to invoke BLT and STRBLT operations, permitting software to set up
instruction buffers and use batch processing as described in the 3D/2D Instruction Processing Section.
Note that these instructions replace the need to do PIO directly to BLT and STRBLT registers, which
speeds up the operation.
148
Datasheet
82815 GMCH
R
4.8.7.1.
Fixed BLT Engine
The rectangular block of data does not change as it is transferred between memory locations. The
allowable memory transfers are between: system memory and display cache, display cache and display
cache, and system memory and system memory. Data to be transferred can consist of regions of memory,
patterns, or solid color fills. A pattern will always be 8x8 pixels wide and may be 8, 16, or 24 bits per
pixel.
The GMCH can expand monochrome data into a color depth of 8, 16, or 24 bits. BLTs can be either
opaque or transparent. Opaque transfers, move the data specified to the destination. Transparent
transfers, compare destination color to source color and write according to the mode of transparency
selected.
Data is horizontally and vertically aligned at the destination. If the destination for the BLT overlaps with
the source memory location, the GMCH can specify which area in memory to begin the BLT transfer.
Use of this BLT engine accelerates the Graphical User Interface (GUI) of Microsoft* Windows.
Hardware is included for all 256 raster operations (Source, Pattern, and Destination) defined by
Microsoft* , including transparent BLT.
4.8.7.2.
Arithmetic Stretch BLT Engine
The stretch BLT function can stretch source data in the X and Y directions to a destination larger or
smaller than the source. Stretch BLT functionality expands a region of memory into a larger or smaller
region using replication and interpolation.
The stretch BLT engine also provides format conversion and data alignment. Through an algorithm
implemented in the mapping engine, object expansion and contraction can occur in the horizontal and
vertical directions.
4.8.8.
Hardware Motion Compensation
The Motion Compensation (MC) process consists of reconstructing a new picture by predicting (either
forward, backward or bidirectionally) the resulting pixel colors from one or more reference pictures. The
GMCH intercepts the DVD pipeline at Motion Compensation and implements Motion Compensation and
subsequent steps in hardware. Performing Motion Compensation in hardware reduces the processor
demand of software-based MPEG-2 decoding and, thus, improves system performance.
The GMCH’s implementation of Hardware Motion Compensation supports a motion smoothing
algorithm. When the system processor is not able to process the MPEG decoding stream in a timely
manner (as can happen in software DVD implementations), the GMCH supports downsampled MPEG
decoding. Downsampling allows for reduced spatial resolution in the MPEG picture while maintaining a
full frame rate; this reduces processor load while maintaining the best video quality possible given the
processor constraints.
Datasheet
149
82815 GMCH
R
4.8.9.
Hardware Cursor
The GMCH allows an unlimited number of cursor patterns to be stored in the display cache or system
memory. Two sets of registers contain the x and y position of the cursor relative to the upper left corner
of the display. The following four cursor modes are provided:
• 32x32 2 bpp AND/XOR 2-plane mode
• 64x64 2 bpp 3-color and transparency mode
• 64x64 2 bpp AND/XOR 2-plane mode
• 64x64 2 bpp 4-color mode
4.8.10.
Overlay Engine
The overlay engine provides a method of merging either video capture data (from an external PCI Video
Capture Adapter) or data delivered by the processor, with the graphics data on the screen. Supported data
formats include YUV 4:2:2, YUV 4:2:0, YUV 4:1:0, YUV 4:1:1, RGB15, and RGB16. The source data
can be mirrored horizontally or vertically or both. Overlay data comes from a buffer located in system
memory. Additionally, the overlay engine can be quadruple buffered to support flipping between
different overlay images. Data can either be transferred into the overlay buffer from the host or from an
external PCI adapter (e.g., DVD hardware or video capture hardware). Buffer swaps can be done by the
host and internally synchronized with the display VBLANK.
The GMCH can accept line widths up to 720 pixels. In addition, overlay source and destination
chromakeying are also supported. Overlay source/destination chromakeying enables blending of the
overlay with the underlying graphics background. Destination color/chroma keying can be used to handle
occluded portions of the overlay window on a pixel-by-pixel basis which is actually an underlay. Source
color/chroma keying is used to handle transparency based on the overlay window on a pixel-by-pixel
basis. This is used when “blue screening” an image in order to overlay the image on a new background
later.
To compensate for overlay color intensity loss due to the non-linear response between display devices,
the overlay engine supports independent gamma correction. In addition, the brightness, saturation, and
contrast of the overlay may be independently varied.
150
Datasheet
82815 GMCH
R
4.8.11.
Display
The display function contains a RAM-based Digital-to-Analog Converter (RAMDAC) that transforms
the digital data from the graphics and video subsystems to analog data for the monitor. The GMCH’s
integrated 230 MHz RAMDAC provides resolution support up to 1600x1200. Circuitry is incorporated
to limit the switching noise generated by the DACs. Three 8-bit DACs provide the R, G, and B signals to
the monitor. Sync signals are properly delayed to match any delays from the D-to-A conversion.
Associated with each DAC is a 256 pallet of colors. The RAMDAC can be operated in either direct or
indexed color mode. In Direct color mode, pixel depths of 15, 16, or 24 bits can be realized. Noninterlaced mode is supported. Gamma correction can be applied to the display output.
The GMCH supports a wide range of resolutions, color depths, and refresh rates via a programmable dot
clock that has a maximum frequency of 230 MHz.
Table 19. Partial List of Display Modes Supported
Resolution
Datasheet
Bits Per Pixel (frequency in Hz)
8-bit Indexed
16-bit
24-bit
320x200
70
70
70
320x240
70
70
70
352x480
70
70
70
352x576
70
70
70
400x300
70
70
70
512x384
70
70
70
640x400
70
70
70
640x480
60,70,72,75,85
60,70,72,75,85
60,70,72,75,85
720x480
75,85
75,85
75,85
720x576
60,75,85
60,75,85
60,75,85
800x600
60,70,72,75,85
60,70,72,75,85
60,70,72,75,85
1024x768
60, 70,72,75,85
60, 70,72,75,85
60, 70,72,75,85
1152x864
60,70,72,75,85
60,70,72,75,85
60,70,72,75,85
1280x720
60,75,85
60,75,85
60,75,85
1280x960
60,75,85
60,75,85
60,75,85
1280x1024
60,70,72,75,85
60,70,72,75,85
60,70,75,85
1600x900
60,75,85
60,75,85
1600x1200
60,70,72,75
151
82815 GMCH
R
4.8.12.
Flat Panel/Digital CRT Interface / 1.8V TV-Out Interface
The GMCH has a dedicated port for Flat Panel support. This port is a 16 bit digital port (4 control bits
and 12 data bits) with a 1.8V interface for high speed signaling. The port is designed to connect to
transmission devices. The port can also be used to interface with an external TV encoder that requires
1.8V signals.
Connecting the GMCH to a flat panel transmitter is demonstrated below. For more details, refer to The
Intel® 815 Chipset Design Guide or the Intel® 815E Chipset Design Guide.
The GMCH supports a variety of Flat Panel display modes and refresh rates that require up to a 65 MHz
dot clock over this interface. Table 20 shows some of the display modes supported by the GMCH.
The GMCH supports scaling for all of the resolutions listed in Table 20. Actual scaling results are
dependant on the third party flat panel transmitter. If the flat panel transmitter does not support scaling,
the resolutions are supported by the GMCH via centering.
Table 21 shows some of the TV-Out modes supported by the GMCH.
Table 20. Partial List of Flat Panel Modes Supported
Resolution
152
Bits Per Pixel (frequency in Hz)
8-bit Indexed
16-bit
24-bit
320x200
60
60
60
320x240
60
60
60
352x480
60
60
60
352x480
60
60
60
352x576
60
60
60
400x300
60
60
60
512x384
60
60
60
640x350
60
60
60
640x400
60
60
60
640x480
60
60
60
720x480
60
60
60
720x576
60
60
60
800x600
60
60
60
1024x768
60
60
60
Datasheet
82815 GMCH
R
Table 21. Partial List of TV-Out Modes Supported
Resolution
1
320x200
320x240
1
352x480
1
352x576
1
400x300
Colors
NTSC
PAL
256
Yes
Yes
16M
Yes
Yes
64k
Yes
Yes
256
Yes
Yes
16M
Yes
Yes
64k
Yes
Yes
256
Yes
Yes
16M
Yes
Yes
64k
Yes
Yes
256
Yes
Yes
16M
Yes
Yes
64k
Yes
Yes
256
Yes
Yes
16M
Yes
Yes
64k
Yes
Yes
640x4001
256
Yes
Yes
640x480
256
Yes
Yes
1
720x480
1
720x576
800x600
64k
Yes
Yes
16M
Yes
Yes
256
Yes
Yes
64k
Yes
Yes
16M
Yes
Yes
256
Yes
Yes
64k
Yes
Yes
16M
Yes
Yes
16
Yes
Yes
256
Yes
Yes
32k
Yes
Yes
64k
Yes
Yes
16M
Yes
Yes
NOTES:
1. These resolutions are supported via centering.
4.8.13.
DDC (Display Data Channel)
DDC is a standard defined by VESA. Its purpose is to allow communication between the host system and
display. Both configuration and control information can be exchanged allowing plug-and-play systems to
be realized. Support for DDC 2B is implemented.
Datasheet
153
82815 GMCH
R
4.9.
System Reset for the GMCH
Refer to the Intel 815 Chipset Design Guide or Intel 815E Chipset Design Guide (Power Sequencing
section) for details.
4.10.
System Clock Description
The Intel 815 chipset family is supported by the CK815 chipset 2DIMM and CK815 chipset 3DIMM
clock synthesizers. For details, refer to the Intel 815 Chipset Design Guide or Intel 815E Chipset
Design Guide.
4.11.
Power Management
4.11.1.
Specifications Supported
The platform is compliant with the following specifications:
154
•
APM Rev 1.2
•
ACPI Rev 1.0
•
PCI Power Management Rev 1.0
•
PC 99 System Design Guide, Rev 1.0
Datasheet
82815 GMCH
R
5.
Pinout and Package Information
5.1.
82815 GMCH Pinout
Figure 10 and Figure 11 show the ball foot print of the GMCH. These figures represent the ballout by
ball number. Table 22 provides an alphabetical signal listing of the ballout.
Datasheet
155
82815 GMCH
R
Figure 10. GMCH Pinout (Top View-Left Side)
1
2
3
4
5
6
7
8
9
10
11
12
13
A
SMD17
SMD49
SMD16
SMD48
SDQM7
SDQM3
SDQM2
SCSB5#
SMAC7#
SMAC5#
SMAA7
SMAA5
SMAA11
B
SMD50
VSUS3.3
SMD56
SMD23
VSUS3.3
SDQM6
SMAA12
VSUS3.3
SCSB4#
SMAC4#
VSUS3.3
SMAA4
SBS0
C
SMD18
SMD25
VSS
SMD55
SMD22
VSS
SCKE5
SCKE4
VSS
SMAC6#
SMAA6
VSS
SMAA9
D
VSS
SMD57
SMD26
SMD24
SMD54
SMD21
SCKE3
SCKE0
SCSB3#
SCSB2#
SBS1
SMAA8
SMAA0
E
SMD51
VSUS3.3
SMD58
SMD27
VSS
SMD53
VSS
SCKE1
SCKE2
VSS
SMAA10
VSS
SCSA4#
F
VSS
SMD19
VSS
SMD59
SMD28
SMD60
SCLK
SCSB1#
SCSB0#
VSUS3.3
VSS
SMAA2
VSS
G
ADS#
SMD52
SMD20
SMD29
SMD61
VSUS3.3
SRCOMP
VSUS3.3
VSS
NC
H
RS2#
VSUS3.3
RESET#
SMD62
VSUS3.3
VSS
VSUS3.3
J
DRDY#
VSS
DBSY#
SMD63
VSS
SMD30
VCCDPLL
K
HIT#
RS0#
HTRDY#
VSS
SMD31
V1.8
VSSDPLL
L
RS1#
VSS
HITM#
HLOCK#
HREQ3#
VSS
VSS
VSS
VSS
M
HREQ0#
HREQ2#
DEFER#
VSS
BPRI#
V1.8
VSS
VSS
VSS
N
HREQ1#
VSS
HREQ4#
BNR#
HA7#
VSS
VSS
VSS
VSS
P
HA4#
HA11#
HA14#
VSS
HA8#
V1.8
VSS
VSS
VSS
R
HA9#
VSS
HA6#
HA3#
HA16#
VSS
VSS
VSS
VSS
T
HA12#
HA5#
HA13#
VSS
HA15#
V1.8
VSS
VSS
VSS
U
HA10#
VSS
HA28#
HA21#
HA25#
GTLREF0
VSS
V
HA31#
HA22#
HA19#
VSS
HA17#
VSS
V1.8
W
HA20#
VSS
HA23#
HA24#
HA30#
V1.8
VSS
Y
HA29#
HA18#
HA27#
VSS
HA26#
VSS
V1.8
VSS
V1.8
VSS
AA
HD0#
VSS
HD6#
HD15#
CPURST#
V1.8
HCLK
V1.8
VSS
GTLREF1
V1.8
VSS
V1.8
AB
HD4#
HD1#
HD5#
VSS
HD23#
HD19#
HD31#
HD34#
HD37#
HD42#
HD41#
HD48#
HD55#
AC
HD8#
VSS
HD17#
HD7#
VSS
HD25#
VSS
HD22#
VSS
HD44#
VSS
HD63#
VSS
AD
HD10#
HD12#
HD13#
HD3#
HD30#
HD16#
HD33#
HD29#
HD43#
HD39#
HD27#
HD47#
HD59#
AE
HD18#
VSS
HD11#
VSS
HD21#
VSS
HD35#
VSS
HD36#
VSS
HD49#
VSS
HD57#
AF
HD14#
HD2#
HD9#
HD20#
HD24#
HD26#
HD32#
HD28#
HD38#
HD45#
HD51#
HD40#
HD52#
1
2
3
4
5
6
7
8
9
10
11
12
13
156
Datasheet
82815 GMCH
R
Figure 11. GMCH Pinout (Top View-Right Side)
14
15
16
17
18
19
20
21
22
23
24
25
26
SMAB7#
SMAB5#
SMAA3
SCSA1#
SDQM4
SMD47
SMD45
SMD42
SMD39
SMD37
SMD35
SMD33
SMD32
A
VSUS3.3
SMAB4#
SMAA1
SCSA5#
SMD12
VSUS3.3
SMD44
SMD41
VSUS3.3
SMD36
SMD34
VSUS3.3
VSS
B
SMAB6#
VSS
SRAS#
SDQM5
VSS
SMD46
SMD43
VSS
SMD38
SMD1
VSS
V1.8
HL10
C
SCSA2#
SCSA0#
SDQM0
SMD15
SCAS#
SMD10
SMD7
SMD40
SMD2
SMD0
HL9
HL8
HL7
D
SCSA3#
VSS
SWE#
VSS
SMD11
SMD9
VSS
SMD4
VSSBA
VCCBA
V1.8
HL6
HL5
E
VSUS3.3
SDQM1
VSS
VSUS3.3
SMD13
SMD8
SMD6
SMD3
HLCLK
V1.8
HL4
VSS
HLPSTRB#
F
VSS
SMD14
VSUS3.3
SMD5
VSS
V1.8
VSS
HL3
HLPSTRB
V1.8
G
HLZCOMP
HLREF
VSS
G_C/BE0#
HL0
HL2
HL1
H
G_AD5
G_AD3
G_AD1
VSS
AGPREF
VSS
GRCOMP
J
VDDQ
VSS
G_AD8
G_AD7
VSS
G_AD2
G_AD0
K
VSS
VSS
VSS
VDDQ
VSS
AD_STB0#
G_AD4
VSS
G_AD6
L
VSS
VSS
VSS
G_AD12
AD_STB0
VDDQ
G_AD10
G_AD9
G_AD11
M
VSS
VSS
VSS
G_C/BE1#
G_AD14
VSS
G_AD13
VDDQ
G_AD15
N
VSS
VSS
VSS
G_TRDY#
LRCLK
G_IRDY#
VSS
G_STOP#
G_DEVSEL#
P
VSS
VSS
VSS
VDDQ
LOCLK
VSS
G_PAR
VSS
G_FRAME#
R
VSS
VSS
VSS
VSS
G_AD17
G_AD19
G_AD21
G_C/BE2#
G_AD16
T
VDDQ
G_AD23
AD_STB1
VDDQ
G_AD18
VDDQ
G_AD20
U
VSS
G_AD25
VSS
AD_STB1#
G_AD22
G_AD24
G_AD26
V
VDDQ
G_AD27
G_AD29
VSS
G_AD28
VSS
G_AD30
W
VSS
V1.8
VSSDA
IWASTE
G_AD31
SBA6
SB_STB
VDDQ
SBA7
G_C/BE3#
Y
VSS
V1.8
VSS
V1.8
DDDA
V1.8
LTVDA
VCCDA
SBA4
VSS
SB_STB#
VSS
SBA5
AA
HD53#
HD56#
V1.8
LTVHSYNC
DDCK
LTVBLANK#
V1.8
LTVCK
SBA0
SBA2
WBF#
SBA1
SBA3
AB
HD58#
VSS
LTVVSYNC
VSS
LTVCLKIN
VSS
LTVDATA8
VSS
V1.8
ST2
ST1
VSS
PIPE#
AC
HD46#
HD60#
LTVDATA0
LTVDATA3
LTVDATA5
V1.8
LTVDATA7
LTVDATA11
RED
IREF
ST0
G_GNT#
RBF#
AD
VSS
HD50#
VSS
LTVDATA2
VSS
LTVCLKOUT0
VSS
LTVDATA10
GREEN
BLUE
DCLKREF
VSSDACA
G_REQ#
AE
HD54#
HD62#
HD61#
LTVDATA1
LTVDATA4
LTVCLKOUT1
LTVDATA6
LTVDATA9
VSYNC
HSYNC
VSSDACA
VCCDACA2
VCCDACA1
AF
14
15
16
17
18
19
20
21
22
23
24
25
26
Datasheet
157
82815 GMCH
R
Table 22. Alphabetical Pin Assignment
Signal Name
Ball #
ADS#
G1
AD_STB0
M22
AD_STB0#
L23
AD_STB1
U22
AD_STB1#
V23
AGPREF
J24
BLUE
AE23
BNR#
N4
BPRI#
M5
CPURST#
AA5
DBSY#
J3
DCLKREF
AE24
DDCK
AB18
DDDA
AA18
DEFER#
M3
DRDY#
J1
G_AD0
K26
G_AD1
J22
G_AD2
K25
G_AD4
L24
G_AD5
J20
G_AD6
L26
G_AD7
K23
G_AD8
K22
G_AD9
M25
G_AD10
M24
G_AD11
M26
G_AD12
M21
G_AD13
N24
G_AD14
N22
G_AD15
N26
G_AD16
T26
G_AD17
T22
G_AD18
U24
G_AD19
T23
158
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
G_AD20
U26
HA10#
U1
HD13#
AD3
G_AD21
T24
HA11#
P2
HD14#
AF1
G_AD22
V24
HA12#
T1
HD15#
AA4
G_AD23
U21
HA13#
T3
HD16#
AD6
G_AD24
V25
HA14#
P3
HD17#
AC3
G_AD25
V21
HA15#
T5
HD18#
AE1
G_AD26
V26
HA16#
R5
HD19#
AB6
G_AD27
W21
HA17#
V5
HD20#
AF4
G_AD28
W24
HA18#
Y2
HD21#
AE5
G_AD29
W22
HA19#
V3
HD22#
AC8
G_AD3
J21
HA20#
W1
HD23#
AB5
G_AD30
W26
HA21#
U4
HD24#
AF5
G_AD31
Y21
HA22#
V2
HD25#
AC6
G_C/BE0#
H23
HA23#
W3
HD26#
AF6
G_C/BE1#
N21
HA24#
W4
HD27#
AD11
G_C/BE2#
T25
HA25#
U5
HD28#
AF8
G_C/BE3#
Y26
HA26#
Y5
HD29#
AD8
G_DEVSEL#
P26
HA27#
Y3
HD30#
AD5
G_FRAME#
R26
HA28#
U3
HD31#
AB7
G_GNT#
AD25
HA29#
Y1
HD32#
AF7
G_IRDY#
P23
HA30#
W5
HD33#
AD7
G_PAR
R24
HA31#
V1
HD34#
AB8
GRCOMP
J26
HCLK
AA7
HD35#
AE7
GREEN
AE22
HD0#
AA1
HD36#
AE9
G_REQ#
AE26
HD1#
AB2
HD37#
AB9
G_STOP#
P25
HD2#
AF2
HD38#
AF9
GTLREF0
U6
HD3#
AD4
HD39#
AD10
GTLREF1
AA10
HD4#
AB1
HD40#
AF12
G_TRDY#
P21
HD5#
AB3
HD41#
AB11
HA3#
R4
HD6#
AA3
HD42#
AB10
HA4#
P1
HD7#
AC4
HD43#
AD9
HA5#
T2
HD8#
AC1
HD44#
AC10
HA6#
R3
HD9#
AF3
HD45#
AF10
HA7#
N5
HD10#
AD1
HD46#
AD14
HA8#
P5
HD11#
AE3
HD47#
AD12
HA9#
R1
HD12#
AD2
HD48#
AB12
Datasheet
82815 GMCH
R
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
HD49#
AE11
HREQ3#
L5
SBA1
AB25
SDQM6
B6
HD50#
AE15
HREQ4#
N3
SBA2
AB23
SDQM7
A5
HD51#
AF11
HSYNC
AF23
SBA3
AB26
SMAA0
D13
HD52#
AF13
HTRDY#
K3
SBA4
AA22
SMAA1
B16
HD53#
AB14
IREF
AD23
SBA5
AA26
SMAA10
E11
HD54#
AF14
IWASTE
Y20
SBA6
Y22
SMAA11
A13
HD55#
AB13
LOCLK
R22
SBA7
Y25
SMAA12
B7
HD56#
AB15
LRCLK
P22
SBS0
B13
SMAA2
F12
HD57#
AE13
LTVBLANK#
AB19
SBS1
D11
SMAA3
A16
HD58#
AC14
LTVCK
AB21
SB_STB
Y23
SMAA4
B12
HD59#
AD13
LTVCLKIN
AC18
SB_STB#
AA24
SMAA5
A12
HD60#
AD15
LTVCLKOUT0
AE19
SCAS#
D18
SMAA6
C11
HD61#
AF16
LTVCLKOUT1
AF19
SCKE0
D8
SMAA7
A11
HD62#
AF15
LTVDA
AA20
SCKE1
E8
SMAA8
D12
HD63#
AC12
LTVDATA0
AD16
SCKE2
E9
SMAA9
C13
HIT#
K1
LTVDATA1
AF17
SCKE3
D7
SMAB4#
B15
HITM#
L3
LTVDATA10
AE21
SCKE4
C8
SMAB5#
A15
HL0
H24
LTVDATA11
AD21
SCKE5
C7
SMAB6#
C14
HL1
H26
LTVDATA2
AE17
SCLK
F7
SMAB7#
A14
HL2
H25
LTVDATA3
AD17
SCSA0#
D15
SMAC4#
B10
HL3
G24
LTVDATA4
AF18
SCSA1#
A17
SMAC5#
A10
HL4
F24
LTVDATA5
AD18
SCSA2#
D14
SMAC6#
C10
HL5
E26
LTVDATA6
AF20
SCSA3#
E14
SMAC7#
A9
HL6
E25
LTVDATA7
AD20
SCSA4#
E13
SMD0
D23
HL7
D26
LTVDATA8
AC20
SCSA5#
B17
SMD1
C23
HL8
D25
LTVDATA9
AF21
SCSB0#
F9
SMD2
D22
HL9
D24
LTVHSYNC
AB17
SCSB1#
F8
SMD3
F21
HL10
C26
LTVVSYNC
AC16
SCSB2#
D10
SMD4
E21
HLCLK
F22
NC
G10
SCSB3#
D9
SMD5
G20
HLOCK#
L4
PIPE#
AC26
SCSB4#
B9
SMD6
F20
HLPSTRB
G25
RBF#
AD26
SCSB5#
A8
SMD7
D20
HLPSTRB#
F26
RED
AD22
SDQM0
D16
SMD8
F19
HLREF
H21
RESET#
H3
SDQM1
F15
SMD9
E19
HLZCOMP
H20
RS0#
K2
SDQM2
A7
SMD10
D19
HREQ0#
M1
RS1#
L1
SDQM3
A6
SMD11
E18
HREQ1#
N1
RS2#
H1
SDQM4
A18
SMD12
B18
HREQ2#
M2
SBA0
AB22
SDQM5
C17
SMD13
F18
Datasheet
159
82815 GMCH
R
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
SMD14
G18
SMD51
E1
V1.8
F23
VSS
R23
SMD15
D17
SMD52
G2
V1.8
G22
VSS
R25
SMD16
A3
SMD53
E6
V1.8
G26
VSS
R6
SMD17
A1
SMD54
D5
V1.8
K6
VSS
T11
SMD18
C1
SMD55
C4
V1.8
M6
VSS
T12
SMD19
F2
SMD56
B3
V1.8
P6
VSS
T13
SMD20
G3
SMD57
D2
V1.8
Y7
VSS
T14
SMD21
D6
SMD58
E3
VCCBA
E23
VSS
T15
SMD22
C5
SMD59
F4
VCCDA
AA21
VSS
T16
SMD23
B4
SMD60
F6
VCCDACA1
AF26
VSS
AC15
SMD24
D4
SMD61
G5
VCCDACA2
AF25
VSS
T21
SMD25
C2
SMD62
H4
VCCDPLL
J7
VSS
T4
SMD26
D3
SMD63
J4
VDDQ
N25
VSS
U2
SMD27
E4
SRAS#
C16
VDDQ
K20
VSS
K24
SMD28
F5
SRCOMP
G7
VDDQ
L21
VSS
U7
SMD29
G4
ST0
AD24
VDDQ
M23
VSS
V22
SMD30
J6
ST1
AC24
VDDQ
U25
VSS
V4
SMD31
K5
ST2
AC23
VDDQ
R21
VSS
V6
SMD32
A26
SWE#
E16
VDDQ
U23
VSS
W2
SMD33
A25
V1.8
C25
VDDQ
W20
VSS
AC17
SMD34
B24
V1.8
T6
VDDQ
Y24
VSS
V20
SMD35
A24
V1.8
V7
VDDQ
U20
VSS
W23
SMD36
B23
V1.8
W6
VSS
F1
VSS
W25
SMD37
A23
V1.8
AA6
VSS
AA12
VSS
W7
SMD38
C22
V1.8
Y9
VSS
AC11
VSS
Y10
SMD39
A22
V1.8
Y18
VSS
P14
VSS
Y17
SMD40
D21
V1.8
AA8
VSS
P15
VSS
P13
SMD41
B21
V1.8
AA11
VSS
P16
VSS
Y4
SMD42
A21
V1.8
AA13
VSS
P4
VSS
Y8
SMD43
C20
V1.8
AA15
VSS
R11
VSS
AC19
SMD44
B20
V1.8
E24
VSS
R12
VSS
AC2
SMD45
A20
V1.8
AA17
VSS
R13
VSS
AC21
SMD46
C19
V1.8
AA19
VSS
R14
VSS
AC25
SMD47
A19
V1.8
AD19
VSS
R15
VSS
AC5
SMD48
A4
V1.8
AC22
VSS
R16
VSS
AC7
SMD49
A2
V1.8
AB16
VSS
AC13
VSS
AA14
SMD50
B1
V1.8
AB20
VSS
R2
VSS
AC9
160
Datasheet
82815 GMCH
R
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
Signal Name
Ball #
VSS
AE10
VSS
E5
VSS
L2
VSSDA
Y19
VSS
AE12
VSS
E7
VSS
L22
VSSDACA
AE25
VSS
AE14
VSS
F11
VSS
L25
VSSDACA
AF24
VSS
AE16
VSS
AA25
VSS
L6
VSSDPLL
K7
VSS
AE18
VSS
F13
VSS
M11
VSUS3.3
F14
VSS
AE2
VSS
F16
VSS
AA23
VSUS3.3
B2
VSS
AE20
VSS
F25
VSS
M12
VSUS3.3
F10
VSS
AE4
VSS
F3
VSS
M13
VSUS3.3
F17
VSS
AE6
VSS
G17
VSS
M14
VSUS3.3
G6
VSS
AA16
VSS
G21
VSS
M15
VSUS3.3
G8
VSS
AE8
VSS
G23
VSS
M16
VSUS3.3
H2
VSS
B26
VSS
G9
VSS
K21
VSUS3.3
H5
VSS
C12
VSS
Y6
VSS
M4
VSUS3.3
H7
VSS
C15
VSS
H22
VSS
N11
VSUS3.3
G19
VSS
C18
VSS
H6
VSS
N12
VSUS3.3
B5
VSS
C21
VSS
J2
VSS
N13
VSUS3.3
B8
VSS
C24
VSS
J23
VSS
AB4
VSUS3.3
B11
VSS
C3
VSS
J25
VSS
N14
VSUS3.3
B14
VSS
C6
VSS
J5
VSS
N15
VSUS3.3
B19
VSS
C9
VSS
K4
VSS
N16
VSUS3.3
B22
VSS
AA2
VSS
L11
VSS
N2
VSUS3.3
B25
VSS
D1
VSS
AA9
VSS
N23
VSUS3.3
E2
VSS
E10
VSS
L12
VSS
P24
VSYNC
AF22
VSS
E12
VSS
L13
VSS
N6
WBF#
AB24
VSS
E15
VSS
L14
VSS
P11
VSS
E17
VSS
L15
VSS
P12
VSS
E20
VSS
L16
VSSBA
E22
Datasheet
161
82815 GMCH
R
5.2.
Package Information
This specification outlines the mechanical dimensions for the GMCH. The package is a 544 ball grid
array (BGA).
Figure 12. GMCH BGA Package Dimensions (Top and Side Views)
D
D1
Pin A1 corner
Pin A1 I.D.
E1 E
45° Chamfer
(4 places)
Top View
A2
A
30°
c
A1
-CSide View
Seating Plane
pkgbga_top&side.vsd
162
Datasheet
82815 GMCH
R
Figure 13. GMCH BGA Package Dimensions (Bottom View)
Pin A1 corner
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
A
b
B
C
D
E
F
G
e
H
J
K
L
M
N
P
R
T
U
V
W
Y
AA
AB
AC
AD
AE
AF
j
e
544 BGA
Bottom View
l
pkgbga_544.vsd
Table 23. Package Dimensions
Symbol
Min
Nominal
Max
Units
A
2.17
2.38
2.59
mm
A1
0.50
0.60
0.70
mm
A2
1.12
1.17
1.22
mm
D
34.80
35.00
35.20
mm
D1
29.75
30.00
30.25
mm
E
34.80
35.00
35.20
mm
E1
29.75
30.00
30.25
mm
e
1.27 (solder ball pitch)
mm
I
1.63 REF.
mm
J
1.63 REF.
mm
M
26 x 26 Matrix
mm
2
b
0.60
0.75
0.90
mm
c
0.55
0.61
0.67
mm
Note
NOTES:
1. All dimensions and tolerances conform to ANSI Y14.5-1982
2. Dimension is measured at maximum solder ball diameter parallel to primary datum (-C-)
3. Primary Datum (-C-) and seating plane are defined by the spherical crowns of the solder balls.
Datasheet
163
82815 GMCH
R
This page is intentionally left blank.
164
Datasheet
82815 GMCH
R
6.
Testability
In the GMCH, the testability for Automated Test Equipment (ATE) board level testing has been changed
from the traditional NAND chain mode to a XOR chain. The GMCH pins are grouped in eight XOR
chains.
An XOR-Tree is a chain of XOR gates each with one of its inputs connected to a GMCH input pin or bidirectional pin (used as an input pin only). The other input of each XOR gate connects to the noninverted output of the previous XOR gate in the chain. The first XOR gate of each chain will have one
pin internally connected tied to Vcc. The output of the last XOR gate is the chain output. Figure 14
shows the GMCH XOR chain implementation.
Figure 14. XOR Tree Implementation
Vcc
XOR
Out
Pin 1
Pin 2
Pin 3
Pin 4
Pin 5
Pin 6
xor.vsd
Datasheet
165
82815 GMCH
R
6.1.
XOR Tree Testability Algorithm Example
XOR tree testing allows users to check, for example, opens and shorts to VCC or GND. An example
algorithm to do this is shown in Table 24.
Table 24. XOR Test Pattern Example
Pin # from Figure 14
Vector
PIN1
PIN2
PIN3
PIN4
PIN5
PIN6
XOROut
1
0
0
0
0
0
0
1
2
1
0
0
0
0
0
0
3
1
1
0
0
0
0
1
4
1
1
1
0
0
0
0
5
1
1
1
1
0
0
1
6
1
1
1
1
1
0
0
7
1
1
1
1
1
1
1
In this example, Vector 1 applies all 0s to the chain inputs. The outputs being non-inverting, will
consistently produce a 1 at the XOR chain output on a good board. One short to Vcc (or open floating to
Vcc) will cause a 0 at the chain output, signaling a defect.
Likewise, applying Vector 7 (all 1s) to chain inputs (given that there are an even number of signals in the
chain), will consistently produce a 1 at the XOR chain output on a good board. One short to Vss (or open
floating to Vss) will cause a 0 at the chain output, signaling a defect. It is important to note that the
number of inputs pulled to 1 will affect the expected chain output value. If the number chain inputs
pulled to 1 is even, then expect 1 at XOR-out; otherwise, if odd, expect 0.
Continuing to Illustrate with the example pattern in Table 24, as the pins are driven to 1 across the chain
in sequence, XOR-out will toggle between 0 and 1. Any break in the toggling sequence (e.g., 1011) will
identify the location of the short or open.
6.1.1.
Test Pattern Consideration for XOR Chains 3 and 4, and 7 and 8
Bi-directional pins HLPSTRB (chain 3) and HLPSTRB# (chain 4), and AGP strobes AD_STB0,
AD_STB1, and SB_STB (chain 7) and AD_STB0#, AD_STB1#, and SB_STB# (chain 8) must always
be complementary to each other. For example, if a 1 is driven on to HLPSTRB, a 0 must be driven on
HLPSTRB# and vice versa. This will need to be considered in applying test patterns to these chains.
166
Datasheet
82815 GMCH
R
6.2.
6.2.1.
XOR Tree Initialization
Chain [1:6] Initialization
On chains [1:6], all that is required to prepare the device for XOR chain testing is to pull SRAS# low
prior to deasserting RESET#. The following sequence will put the GMCH into XOR testability mode:
1. Deassert RESET# high and assert SRAS# low
2. Assert RESET# low; maintain SRAS# low
3. Deassert RESET# high; maintain SRAS# low
4. RESET# must be maintained high for the duration of testing.
No external clocking of the GMCH is required for testing these chains.
6.2.2.
Chain [7:8] Initialization
On chains[7:8], all that is required to prepare the device for XOR chain testing is to pull SMAA2 low
prior to deasserting RESET#, then set LTVCK high and LTVDATA[11:6] to [101101] (1 means high
and 0 means low), follow other LTVCK high and LTVDATA[11:6] to [100011]. The following
sequence puts the GMCH into XOR testability mode for Chain[7:8] only:
1. Deassert RESET# high and assert SMAA2 low
2. Assert RESET# low; maintain SMAA2 low
3. Deassert RESET# high; maintain SMAA2 low
4. Deassert LTVCK low and assert LTVDATA[11:6] to [101101]
5. Assert LTVCK high; maintain LTVDATA[11:6] to [101101]
6. Deassert LTVCK low; maintain LTVDATA[11:6] to [101101]
7. Deassert LTVCK low and assert LTVDATA[11:6] to [100011]
8. Assert LTVCK high; maintain LTVDATA[11:6] to [100011]
9. Deassert LTVCK low; maintain LTVDATA[11:6] to [100011]
10. RESET# must be maintained high for the duration of testing
No external clocking of the GMCH is required for testing these chains.
Datasheet
167
82815 GMCH
R
6.3.
XOR Chain
This is the primary test mode for checking the IO buffer connectivity. There are a total of 8 XOR chains
each containing less than 65 XOR gates. The XOR gates are physically located in the IO buffers. This
test mode can be invoked with the use of reset straps.
Table 25 XOR Chain 1
Pin Name
35 Inputs
Ball
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
DEFER#
M3
1.5V
HD12#
AD2
1.5V
HD30#
AD5
1.5V
HD0#
AA1
1.5V
HD18#
AE1
1.5V
HD9#
AF3
1.5V
HA30#
W5
1.5V
HD7#
AC4
1.5V
HD16#
AD6
1.5V
HD4#
AB1
1.5V
HD13#
AD3
1.5V
HD21#
AE5
1.5V
HD6#
AA3
1.5V
HD23#
AB5
1.5V
HD20#
AF4
1.5V
HA26#
AF6
1.5V
HD14#
AF1
1.5V
HD24#
AF5
1.5V
HD1#
AB2
1.5V
HD2#
AF2
1.5V
HD22#
AC8
1.5V
HD15#
AA4
1.5V
HD19#
AB6
1.5V
HD26#
AF6
1.5V
HD8#
AC1
1.5V
HD3#
AD4
1.5V
HD29#
AD8
1.5V
HD5#
AB3
1.5V
HD11#
AE3
1.5V
HD28#
AF8
1.5V
HD10#
AD1
1.5V
HD31#
AB7
1.5V
HD27#
AD11
1.5V
HD17#
AC3
1.5V
HD25#
AC6
1.5V
Pin Name
Ball
Table 26 XOR Chain 2
Pin Name
168
Voltage
Output: SMAA5 (A12)
33 Inputs
Ball
Voltage
Output: SMAA2 (F12)
Pin Name
Ball
Voltage
Voltage
BPRI#
M5
1.5V
HD38#
AF9
1.5V
HD55#
AB13
1.5V
HD34#
AB8
1.5V
HD45#
AF10
1.5V
HD52#
AF13
1.5V
HD33#
AD7
1.5V
HD41#
AB11
1.5V
HD58#
AC14
1.5V
HD37#
AB9
1.5V
HD49#
AE11
1.5V
HD46#
AD14
1.5V
HD35#
AE7
1.5V
HD63#
AC12
1.5V
HD54#
AF14
1.5V
HD32#
AF7
1.5V
HD51#
AF11
1.5V
HD53#
AB14
1.5V
HD43#
AD9
1.5V
HD47#
AD12
1.5V
HD62#
AF15
1.5V
HD44#
AC10
1.5V
HD48#
AB12
1.5V
HD50#
AE15
1.5V
HD36#
AE9
1.5V
HD40#
AF12
1.5V
HD60#
AD15
1.5V
HD42#
AB10
1.5V
HD59#
AD13
1.5V
HD56#
AB15
1.5V
HD39#
AD10
1.5V
HD57#
AE13
1.5V
HD61#
AF16
1.5V
Datasheet
82815 GMCH
R
Table 27 XOR Chain 3
Pin Name
38 Inputs
Ball
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
HL1
H26
1.8V
HA7#
N5
1.5V
HA24#
W4
1.5V
HL0
H24
1.8V
HA14#
P3
1.5V
HA27#
Y3
1.5V
HLPSTRB
G25
1.8V
HA8#
P5
1.5V
CPURST#
AA5
1.5V
HL5
E26
1.8V
HA3#
R4
1.5V
LTVHSYNC
AB17
1.8V
HL6
E25
1.8V
HA9#
R1
1.5V
LTVCLKIN
AC18
1.8V
HL8
D25
1.8V
HA12#
T1
1.5V
LTVDATA6
AF20
1.8V
ADS#
G1
1.5V
HA13#
T3
1.5V
LTVDATA9
AF21
1.8V
HTRDY#
K3
1.5V
HA10#
U1
1.5V
LTVDATA7
AD20
1.8V
DRDY#
J1
1.5V
HA21#
AE5
1.5V
LTVBLANK#
AB19
1.8V
RS0#
K2
1.5V
HA22#
AC8
1.5V
LTVDATA10
AE21
1.8V
HIT#
K1
1.5V
HA19#
AB6
1.5V
LTVDATA8
AC20
1.8V
HREQ2#
M2
1.5V
HA23#
W3
1.5V
LTVDATA11
AD21
1.8V
HREQ0#
M1
1.5V
HA17#
V5
1.5V
Table 28 XOR Chain 4
Pin Name
Datasheet
Voltage
Output: SMAA0 (D13)
36 Inputs
Ball
Voltage
Output: SMAA9 (D13)
Pin Name
Ball
Voltage
Pin Name
HL2
H25
1.8V
HREQ4#
N3
1.5V
HA18#
HL3
G24
1.8V
HREQ1#
N1
1.5V
HLPSTRB#
F26
1.8V
HA11#
P2
HL4
F24
1.8V
HA4#
HL7
D26
1.8V
HL10
C26
DBSY#
Ball
Voltage
Y2
1.5V
LTVDATA0
AD16
1.8V
1.5V
LTVVSYNC
AC16
1.8V
P1
1.5V
LTVDATA1
AF17
1.8V
HA6#
R3
1.5V
LTVDATA2
AE17
1.8V
1.8V
HA16#
AD6
1.5V
LTVDATA3
AD17
1.8V
J3
1.5V
HA5#
T2
1.5V
LTVDATA4
AF18
1.8V
RS2#
H1
1.5V
HA15#
T5
1.5V
LTVDATA5
AD18
1.8V
HLOCK#
L4
1.5V
HA28#
U3
1.5V
LTVCLKOUT1
AF19
1.8V
HREQ3#
L5
1.5V
HA31#
V1
1.5V
LTVCLKOUT0
AE19
1.8V
HITM#
L3
1.5V
HA25#
U5
1.5V
HA20#
W1
1.5V
RS1#
L1
1.5V
BNR#
N4
1.5V
HA29#
Y1
1.5V
169
82815 GMCH
R
Table 29 XOR Chain 5
Pin Name
170
56 Inputs
Ball
Voltage
Output: SMD31 (K5)
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
SBS1
D11
3.3V
SCKE5
C7
3.3V
SMD59
F4
3.3V
SMAA10
E11
3.3V
SCKE3
D7
3.3V
SMD50
B1
3.3V
SMAC5#
A10
3.3V
SMD39
A22
3.3V
SMD58
E3
3.3V
SMAC4#
B10
3.3V
SDQM7
A5
3.3V
SMD57
D2
3.3V
SMAC6#
C10
3.3V
SMD53
E6
3.3V
SMD62
H4
3.3V
SMAC7#
A9
3.3V
SMD54
D5
3.3V
SMD63
J4
3.3V
SMAA12
B7
3.3V
SMD48
A4
3.3V
SMD51
E1
3.3V
SDQM6
B6
3.3V
SMD55
C4
3.3V
SMD52
G2
3.3V
SCKE4
C8
3.3V
SMD60
F6
3.3V
SCSA2#
D14
3.3V
SMAB4#
B15
3.3V
SMD56
B3
3.3V
SMAB6#
C14
3.3V
SMD36
B23
3.3V
SMD49
A2
3.3V
SMAB5#
A15
3.3V
SMD35
A24
3.3V
SMD61
G5
3.3V
SMAB7#
A14
3.3V
VSYNC
AF22
3.3V
SDQM5
C17
3.3V
SCSA4#
E13
3.3V
SMD32
A26
3.3V
SCSA5#
B17
3.3V
SMD37
A23
3.3V
SMD41
B21
3.3V
SDQM4
A18
3.3V
SMD34
B24
3.3V
SMD42
A21
3.3V
SMD46
C19
3.3V
SMD33
A25
3.3V
SCSA0#
D15
3.3V
SMD45
A20
3.3V
SMD40
D21
3.3V
SCSA1#
A17
3.3V
SMD47
A19
3.3V
SMD38
C22
3.3V
SCSA3#
E14
3.3V
SMD43
C20
3.3V
Datasheet
82815 GMCH
R
Table 30 XOR Chain 6
Pin Name
Ball
Voltage
Output: SMAA11 (A13)
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
LOCLK
R22
3.3V
SMD11
E18
3.3V
SRAS#
C16
3.3V
LRCLK
P22
3.3V
SMD10
D19
3.3V
SMAA1
B16
3.3V
SMD3
F21
3.3V
SMD13
F18
3.3V
SMAA3
A16
3.3V
SMD0
D23
3.3V
SMD44
B20
3.3V
SBS0
B13
3.3V
SMD5
G20
3.3V
SCAS#
D18
3.3V
LTVCK
AB21
3.3V
SMD6
F20
3.3V
SMD15
D17
3.3V
LTVDA
AA20
3.3V
SMD4
E21
3.3V
SWE#
E16
3.3V
DDCK
AB18
3.3V
SMD2
D22
3.3V
SMD12
B18
3.3V
DDDA
AA18
3.3V
SMD1
C23
3.3V
SDQM1
F15
3.3V
HSYNC
AF23
3.3V
SMD8
F19
3.3V
SDQM0
D16
3.3V
SMD19
F2
3.3V
SMD9
E19
3.3V
SMD27
E4
3.3V
SMD30
J6
3.3V
SMD7
D20
3.3V
SMD25
C2
3.3V
SMD20
G3
3.3V
SMD14
G18
3.3V
SMD29
G4
3.3V
SCSB1#
F8
3.3V
SCSB4#
B9
3.3V
SMD24
D4
3.3V
SDQM2
A7
3.3V
SCSB3#
D9
3.3V
SMD28
F5
3.3V
SCKE0
D8
3.3V
SCSB5#
A8
3.3V
SMD26
D3
3.3V
SCKE1
E8
3.3V
SCKE2
E9
3.3V
SMD17
A1
3.3V
SDQM3
A6
3.3V
SMAA7
A11
3.3V
SMD18
C1
3.3V
SMD21
D6
3.3V
SMAA6
C11
3.3V
SMD23
B4
3.3V
SMD22
C5
3.3V
SCSB2#
D10
3.3V
SMD16
A3
3.3V
SCSB0#
F9
3.3V
Table 31 XOR Chain 7
Pin Name
33 Inputs
Ball
Voltage
Output: SMAA8 (D12)
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
ST2
AC23
Vddq
GAD25
V21
Vddq
G_STOP#
P25
Vddq
ST0
AD24
Vddq
G_AD28
W24
Vddq
G_AD14
N22
Vddq
G_GNT#
AD25
Vddq
G_AD22
V24
Vddq
G_AD13
N24
Vddq
RBF#
AD26
Vddq
AD_STB1
U22
Vddq
G_AD11
M26
Vddq
PIPE#
AC26
Vddq
G_AD26
V26
Vddq
G_AD9
M25
Vddq
SBA4
AA22
Vddq
G_AD17
T22
Vddq
AD_STB0
M22
Vddq
Y23
Vddq
G_AD21
T24
Vddq
G_AD4
L24
Vddq
SBA3
AB26
Vddq
G_AD18
U24
Vddq
G_AD7
K23
Vddq
SBA5
AA26
Vddq
G_C/BE2#
T25
Vddq
G_AD1
J22
Vddq
G_C/BE3#
Y26
Vddq
G_PAR
R24
Vddq
G_AD3
J21
Vddq
G_AD27
W21
Vddq
G_TRDY#
P21
Vddq
G_C/BE0#
H23
Vddq
SB_STB
Datasheet
60 Inputs
171
82815 GMCH
R
Table 32 XOR Chain 8
Pin Name
6.4.
31 Inputs
Ball
Voltage
Output: SMAA4 (B12)
Pin Name
Ball
Voltage
Pin Name
Ball
Voltage
ST1
AC24
Vddq
G_AD30
W26
Vddq
G_AD23
U21
Vddq
G_REQ#
AE26
Vddq
AD_STB1#
V23
Vddq
G_AD12
M21
Vddq
WBF#
AB24
Vddq
G_AD24
V25
Vddq
G_AD10
M24
Vddq
SBA0
AB22
Vddq
G_AD19
T23
Vddq
G_AD8
K22
Vddq
SBA2
AB23
Vddq
G_AD20
U26
Vddq
AD_STB0#
L23
Vddq
SBA6
Y22
Vddq
G_AD16
T26
Vddq
G_AD6
L26
Vddq
SB_STB#
AA24
Vddq
G_IRDY#
P23
Vddq
G_AD2
K25
Vddq
SBA1
AB25
Vddq
G_FRAME#
R26
Vddq
G_AD5
J20
Vddq
SBA7
Y25
Vddq
G_DEVSEL#
P26
Vddq
G_AD0
K26
Vddq
G_AD29
W22
Vddq
G_C/BE1#
N21
Vddq
G_AD31
Y21
Vddq
G_AD15
N26
Vddq
All Z
To apply vectors to XOR chains on a system board, other chips on the board must be tri-stated to allow
for this vector application. This is a feature that enables all GMCH outputs to be tristated when the I/O
Controller Hub is in the XOR chain mode. This mode can also be activate using the assigned reset strap.
Tri-state GMCH Outputs
When testing other devices in the system, the GMCH outputs can be tri-stated. To tri-state these outputs
pull the SMAA10 pin low (GND) prior to deasserting RESET#. The following sequence will put the
GMCH into tri-state mode:
1. 1. Deassert RESET# high and SMAA10 low
2. 2. Assert RESET# low; maintain SMAA10 low
3. 3. Deassert RESET# high; maintain SMAA10 low
4. 4. RESET# must be maintained high for the duration of testing.
No external clocking of the GMCH is required.
172
Datasheet