The AMD Instinct MI100 "Arcturus" GPU has had multiple revisions with different part numbers. Very little information is available online about them.
MI100 D34315 Rev 01, certified 2020:
P/N: D34315
VBIOS P/N: 113-D3431500-101, 113-D3431500-100
rocm-smi "Device Rev": 0x01
# amdvbflash -biosfileinfo gpu1.rom
AMDVBFLASH version 4.79 EXTERNAL, Copyright (c) 2020 Advanced Micro Devices, Inc.
Product Name is : MI100 D34315 A2 XL 300W 32GB 1200m
Device ID is : 738C
Bios Version : 000.000.000.000.016113
Bios P/N is : 113-D3431500-101
Bios SSID : 0C34
Bios SVID : 1002
Bios Date is : 11/24/20 22:21
RRA Certification - 2020-08-20
MI100 D34324, Also Rev 01 ??, certified 2021
P/N: D34324
VBIOS P/N: 113-D3432400-100
rocm-smi "Device Rev": 0x01
# amdvbflash -biosfileinfo gpu0.rom
AMDVBFLASH version 4.79 EXTERNAL, Copyright (c) 2020 Advanced Micro Devices, Inc.
Product Name is : MI100 D34324 A2 XL 300W 32GB 1200m
Device ID is : 738C
Bios Version : 000.000.000.000.017958
Bios P/N is : 113-D3432400-100
Bios SSID : 0C34
Bios SVID : 1002
Bios Date is : 12/07/21 23:20
RRA certification - 2021-12-20
Hardware Differences
My D34324 MI100s have some visible differences to their board layout near INJECTION_SW. A TPS2121 power mux has been removed and nearby some 0 ohm linking resistors have been added. I don't know what this power mux is for.
As there were visible hardware changes, I have not tried cross flashing. It might work, but I'm not tempted enough to find out since I have not noticed any issues unique to the D34315 cards.
System Management Command Output
Output from various system management commands on my test system.
# amdvbflash -i
AMDVBFLASH version 4.79 EXTERNAL, Copyright (c) 2020 Advanced Micro Devices, Inc.
adapter seg bn dn dID asic flash romsize test bios p/n
======= ==== == == ==== =============== ============== ======= ==== ================
0 0000 05 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3432400-100
1 0000 45 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3431500-101
2 0000 48 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3431500-101
3 0000 8B 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3432400-100
4 0000 C5 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3431500-101
5 0000 C8 00 738C MI100(Slave) GD25Q80C 100000 pass 113-D3431500-101
rocm-smi info:
$ nix run .#rocmPackages.rocm-smi -- --showhw --alldevices -i -v
====================================== ROCm System Management Interface ======================================
=========================================== Concise Hardware Info ============================================
GPU NODE DID GUID GFX VER GFX RAS SDMA RAS UMC RAS VBIOS BUS PARTITION ID
0 8 0x738c 45705 gfx9008 ENABLED ENABLED ENABLED 113-D3432400-100 0000:05:00.0 0
1 9 0x738c 17080 gfx9008 ENABLED ENABLED ENABLED 113-D3431500-101 0000:45:00.0 0
2 10 0x738c 55164 gfx9008 ENABLED ENABLED ENABLED 113-D3431500-101 0000:48:00.0 0
3 11 0x738c 30316 gfx9008 ENABLED ENABLED ENABLED 113-D3432400-100 0000:8B:00.0 0
4 12 0x738c 58073 gfx9008 ENABLED ENABLED ENABLED 113-D3431500-101 0000:C5:00.0 0
5 13 0x738c 30493 gfx9008 ENABLED ENABLED ENABLED 113-D3431500-101 0000:C8:00.0 0
==============================================================================================================
============================================ End of ROCm SMI Log =============================================
=========================================== ID ===========================================
GPU[0] : Device Name: 0x1002
GPU[0] : Device ID: 0x738c
GPU[0] : Device Rev: 0x01
GPU[0] : Subsystem ID: 0x1002
GPU[0] : GUID: 45705
GPU[1] : Device Name: 0x1002
GPU[1] : Device ID: 0x738c
GPU[1] : Device Rev: 0x01
GPU[1] : Subsystem ID: 0x1002
GPU[1] : GUID: 17080
GPU[2] : Device Name: 0x1002
GPU[2] : Device ID: 0x738c
GPU[2] : Device Rev: 0x01
GPU[2] : Subsystem ID: 0x1002
GPU[2] : GUID: 55164
GPU[3] : Device Name: 0x1002
GPU[3] : Device ID: 0x738c
GPU[3] : Device Rev: 0x01
GPU[3] : Subsystem ID: 0x1002
GPU[3] : GUID: 30316
GPU[4] : Device Name: 0x1002
GPU[4] : Device ID: 0x738c
GPU[4] : Device Rev: 0x01
GPU[4] : Subsystem ID: 0x1002
GPU[4] : GUID: 58073
GPU[5] : Device Name: 0x1002
GPU[5] : Device ID: 0x738c
GPU[5] : Device Rev: 0x01
GPU[5] : Subsystem ID: 0x1002
GPU[5] : GUID: 30493
==========================================================================================
========================================= VBIOS ==========================================
GPU[0] : VBIOS version: 113-D3432400-100
GPU[1] : VBIOS version: 113-D3431500-101
GPU[2] : VBIOS version: 113-D3431500-101
GPU[3] : VBIOS version: 113-D3432400-100
GPU[4] : VBIOS version: 113-D3431500-101
GPU[5] : VBIOS version: 113-D3431500-101
==========================================================================================
dmesg while amdgpu enables the first two GPUs.
$ uname -a
Linux tsukiakari-nixos 6.11.4 #1-NixOS SMP PREEMPT_DYNAMIC Thu Oct 17 13:27:02 UTC 2024 x86_64 GNU/Linux
# dmesg | grep amdgpu | head -n115
[ 22.170939] amdgpu 0000:05:00.0: enabling device (0000 -> 0003)
[ 22.185546] amdgpu 0000:05:00.0: amdgpu: Fetched VBIOS from ROM BAR
[ 22.185552] amdgpu: ATOM BIOS: 113-D3432400-100
[ 22.194004] amdgpu 0000:05:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[ 22.194134] amdgpu 0000:05:00.0: amdgpu: MEM ECC is active.
[ 22.194136] amdgpu 0000:05:00.0: amdgpu: SRAM ECC is active.
[ 22.194154] amdgpu 0000:05:00.0: amdgpu: RAS INFO: ras initialized successfully, hardware ability[67f7f] ras_mask[67f7f]
[ 22.194258] amdgpu 0000:05:00.0: amdgpu: VRAM: 32752M 0x0000008000000000 - 0x00000087FEFFFFFF (32752M used)
[ 22.194262] amdgpu 0000:05:00.0: amdgpu: GART: 512M 0x0000000000000000 - 0x000000001FFFFFFF
[ 22.194534] [drm] amdgpu: 32752M of VRAM memory ready
[ 22.194537] [drm] amdgpu: 257869M of GTT memory ready.
[ 22.253202] amdgpu 0000:05:00.0: amdgpu: reserve 0x400000 from 0x87fe800000 for PSP TMR
[ 22.320931] amdgpu 0000:05:00.0: amdgpu: RAP: optional rap ta ucode is not available
[ 22.320988] amdgpu 0000:05:00.0: amdgpu: use vbios provided pptable
[ 22.320991] amdgpu 0000:05:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.6
[ 22.320994] amdgpu 0000:05:00.0: amdgpu: PMFW based fan control disabled
[ 22.323882] amdgpu 0000:05:00.0: amdgpu: SMU is initialized successfully!
[ 22.552146] amdgpu: HMM registered 32752MB device memory
[ 22.555235] kfd kfd: amdgpu: Allocated 3969056 bytes on gart
[ 22.555279] kfd kfd: amdgpu: Total number of KFD nodes to be created: 1
[ 22.555715] amdgpu: Virtual CRAT table created for GPU
[ 22.556830] amdgpu: Topology: Add dGPU node [0x738c:0x1002]
[ 22.556834] kfd kfd: amdgpu: added device 1002:738c
[ 22.556859] amdgpu 0000:05:00.0: amdgpu: SE 8, SH per SE 1, CU per SH 16, active_cu_number 120
[ 22.556866] amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 0 on hub 0
[ 22.556869] amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 1 on hub 0
[ 22.556872] amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 4 on hub 0
[ 22.556874] amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 5 on hub 0
[ 22.556877] amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 6 on hub 0
[ 22.556879] amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 7 on hub 0
[ 22.556881] amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 8 on hub 0
[ 22.556883] amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 9 on hub 0
[ 22.556886] amdgpu 0000:05:00.0: amdgpu: ring kiq_0.2.1.0 uses VM inv eng 10 on hub 0
[ 22.556888] amdgpu 0000:05:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 8
[ 22.556890] amdgpu 0000:05:00.0: amdgpu: ring sdma1 uses VM inv eng 1 on hub 8
[ 22.556893] amdgpu 0000:05:00.0: amdgpu: ring sdma2 uses VM inv eng 4 on hub 8
[ 22.556895] amdgpu 0000:05:00.0: amdgpu: ring sdma3 uses VM inv eng 5 on hub 8
[ 22.556897] amdgpu 0000:05:00.0: amdgpu: ring sdma4 uses VM inv eng 6 on hub 8
[ 22.556899] amdgpu 0000:05:00.0: amdgpu: ring sdma5 uses VM inv eng 0 on hub 12
[ 22.556902] amdgpu 0000:05:00.0: amdgpu: ring sdma6 uses VM inv eng 1 on hub 12
[ 22.556904] amdgpu 0000:05:00.0: amdgpu: ring sdma7 uses VM inv eng 4 on hub 12
[ 22.556906] amdgpu 0000:05:00.0: amdgpu: ring vcn_dec_0 uses VM inv eng 5 on hub 12
[ 22.556909] amdgpu 0000:05:00.0: amdgpu: ring vcn_enc_0.0 uses VM inv eng 6 on hub 12
[ 22.556911] amdgpu 0000:05:00.0: amdgpu: ring vcn_enc_0.1 uses VM inv eng 7 on hub 12
[ 22.556913] amdgpu 0000:05:00.0: amdgpu: ring vcn_dec_1 uses VM inv eng 8 on hub 12
[ 22.556916] amdgpu 0000:05:00.0: amdgpu: ring vcn_enc_1.0 uses VM inv eng 9 on hub 12
[ 22.556918] amdgpu 0000:05:00.0: amdgpu: ring vcn_enc_1.1 uses VM inv eng 10 on hub 12
[ 22.556920] amdgpu 0000:05:00.0: amdgpu: ring jpeg_dec_0 uses VM inv eng 11 on hub 12
[ 22.556923] amdgpu 0000:05:00.0: amdgpu: ring jpeg_dec_1 uses VM inv eng 12 on hub 12
[ 22.570774] amdgpu 0000:05:00.0: amdgpu: overdrive feature is not supported
[ 22.570892] amdgpu: Detected AMDGPU 6 Perf Events.
[ 22.573492] amdgpu 0000:05:00.0: amdgpu: Runtime PM not available
[ 22.573688] [drm] Initialized amdgpu 3.59.0 for 0000:05:00.0 on minor 1
[ 22.637006] amdgpu 0000:45:00.0: enabling device (0000 -> 0003)
[ 22.647678] amdgpu 0000:45:00.0: amdgpu: Fetched VBIOS from ROM BAR
[ 22.647683] amdgpu: ATOM BIOS: 113-D3431500-101
[ 22.649400] amdgpu 0000:45:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[ 22.649527] amdgpu 0000:45:00.0: amdgpu: MEM ECC is active.
[ 22.649530] amdgpu 0000:45:00.0: amdgpu: SRAM ECC is active.
[ 22.649553] amdgpu 0000:45:00.0: amdgpu: RAS INFO: ras initialized successfully, hardware ability[67f7f] ras_mask[67f7f]
[ 22.649646] amdgpu 0000:45:00.0: amdgpu: VRAM: 32752M 0x0000008000000000 - 0x00000087FEFFFFFF (32752M used)
[ 22.649650] amdgpu 0000:45:00.0: amdgpu: GART: 512M 0x0000000000000000 - 0x000000001FFFFFFF
[ 22.649942] [drm] amdgpu: 32752M of VRAM memory ready
[ 22.649945] [drm] amdgpu: 257869M of GTT memory ready.
[ 22.708825] amdgpu 0000:45:00.0: amdgpu: reserve 0x400000 from 0x87fe800000 for PSP TMR
[ 22.778034] amdgpu 0000:45:00.0: amdgpu: RAP: optional rap ta ucode is not available
[ 22.778090] amdgpu 0000:45:00.0: amdgpu: use vbios provided pptable
[ 22.778092] amdgpu 0000:45:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.6
[ 22.781029] amdgpu 0000:45:00.0: amdgpu: SMU is initialized successfully!
[ 23.014695] amdgpu: HMM registered 32752MB device memory
[ 23.017872] kfd kfd: amdgpu: Allocated 3969056 bytes on gart
[ 23.017918] kfd kfd: amdgpu: Total number of KFD nodes to be created: 1
[ 23.018425] amdgpu: Virtual CRAT table created for GPU
[ 23.020488] amdgpu: Topology: Add dGPU node [0x738c:0x1002]
[ 23.020493] kfd kfd: amdgpu: added device 1002:738c
[ 23.020519] amdgpu 0000:45:00.0: amdgpu: SE 8, SH per SE 1, CU per SH 16, active_cu_number 120
[ 23.020526] amdgpu 0000:45:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 0 on hub 0
[ 23.020529] amdgpu 0000:45:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 1 on hub 0
[ 23.020531] amdgpu 0000:45:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 4 on hub 0
[ 23.020534] amdgpu 0000:45:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 5 on hub 0
[ 23.020536] amdgpu 0000:45:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 6 on hub 0
[ 23.020538] amdgpu 0000:45:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 7 on hub 0
[ 23.020540] amdgpu 0000:45:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 8 on hub 0
[ 23.020543] amdgpu 0000:45:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 9 on hub 0
[ 23.020545] amdgpu 0000:45:00.0: amdgpu: ring kiq_0.2.1.0 uses VM inv eng 10 on hub 0
[ 23.020547] amdgpu 0000:45:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 8
[ 23.020550] amdgpu 0000:45:00.0: amdgpu: ring sdma1 uses VM inv eng 1 on hub 8
[ 23.020552] amdgpu 0000:45:00.0: amdgpu: ring sdma2 uses VM inv eng 4 on hub 8
[ 23.020554] amdgpu 0000:45:00.0: amdgpu: ring sdma3 uses VM inv eng 5 on hub 8
[ 23.020556] amdgpu 0000:45:00.0: amdgpu: ring sdma4 uses VM inv eng 6 on hub 8
[ 23.020559] amdgpu 0000:45:00.0: amdgpu: ring sdma5 uses VM inv eng 0 on hub 12
[ 23.020561] amdgpu 0000:45:00.0: amdgpu: ring sdma6 uses VM inv eng 1 on hub 12
[ 23.020563] amdgpu 0000:45:00.0: amdgpu: ring sdma7 uses VM inv eng 4 on hub 12
[ 23.020566] amdgpu 0000:45:00.0: amdgpu: ring vcn_dec_0 uses VM inv eng 5 on hub 12
[ 23.020568] amdgpu 0000:45:00.0: amdgpu: ring vcn_enc_0.0 uses VM inv eng 6 on hub 12
[ 23.020570] amdgpu 0000:45:00.0: amdgpu: ring vcn_enc_0.1 uses VM inv eng 7 on hub 12
[ 23.020573] amdgpu 0000:45:00.0: amdgpu: ring vcn_dec_1 uses VM inv eng 8 on hub 12
[ 23.020575] amdgpu 0000:45:00.0: amdgpu: ring vcn_enc_1.0 uses VM inv eng 9 on hub 12
[ 23.020578] amdgpu 0000:45:00.0: amdgpu: ring vcn_enc_1.1 uses VM inv eng 10 on hub 12
[ 23.020580] amdgpu 0000:45:00.0: amdgpu: ring jpeg_dec_0 uses VM inv eng 11 on hub 12
[ 23.020582] amdgpu 0000:45:00.0: amdgpu: ring jpeg_dec_1 uses VM inv eng 12 on hub 12
[ 23.034372] amdgpu 0000:45:00.0: amdgpu: overdrive feature is not supported
[ 23.034505] amdgpu: Detected AMDGPU 6 Perf Events.
[ 23.037097] amdgpu 0000:45:00.0: amdgpu: Runtime PM not available
[ 23.040319] [drm] Initialized amdgpu 3.59.0 for 0000:45:00.0 on minor 2