Radeon Pro v340 Drivers

I have been tinkering with a Radeon Pro v340 off and on for a while because I happened on a couple of them way back, however I have never been able to get it to be recognized. I thought it might be related to resizable bar issues and a peculiar motherboard so I put it back and forgot about it. Recently, I tried it again on 3 different systems (Epyc ROMED-8 2T - PopOS 24, Xeon Workstation PopOS 22, Gaming PC - Ubuntu 22 and again on the Workstation Ubuntu 24). I know the 6.19 kernel resurrects some old cards with a new amdgpu rewrite so I even tried that and still nothing. I have tried it with ROCm 5.7, 6.3, 6.4, 7, etc). It always fails with: `38724.150766] [drm] PSP loading VCE firmware` `[38724.302690] amdgpu 0000:09:00.0: amdgpu: reserve 0x400000 from 0x87fe000000 for PSP TMR` `[38724.382050] amdgpu 0000:09:00.0: amdgpu: memory partition mode query is not supported` `[38724.386037] amdgpu 0000:09:00.0: amdgpu: RAP: optional rap ta ucode is not available` `[38724.388798] amdgpu 0000:09:00.0: amdgpu: [drm] Display Core v3.2.351 initialized on DCE 12.1` `[38724.392198] snd_hda_intel 0000:09:00.1: bound 0000:09:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu])` `[38724.544921] amdgpu 0000:09:00.0: amdgpu: kiq ring mec 2 pipe 1 q 0` `[38724.903628] amdgpu: HMM registered 32752MB device memory` `[38724.904539] kfd kfd: amdgpu: Allocated 3969056 bytes on gart` `[38724.904568] kfd kfd: amdgpu: Total number of KFD nodes to be created: 1` `[38724.904572] amdgpu: [powerplay] [MemMclks]: memclk dpm not enabled!` `[38724.904713] amdgpu: Virtual CRAT table created for GPU` `[38724.904852] amdgpu: [powerplay] [MemMclks]: memclk dpm not enabled!` `[38724.904856] amdgpu: Topology: Add dGPU node [0x66a3:0x1002]` `[38724.904857] kfd kfd: amdgpu: added device 1002:66a3` `[38724.905569] [drm:smu_v11_0_i2c_xfer [amdgpu]] *ERROR* Received I2C_NAK_7B_ADDR_NOACK !!!` `[38724.906073] [drm:smu_v11_0_i2c_xfer [amdgpu]] *ERROR* WriteI2CData() - I2C error occurred :1` `[38724.906573] amdgpu 0000:09:00.0: amdgpu: Couldn't read the IPMI Common Header: -5` `[38724.906587] amdgpu 0000:09:00.0: amdgpu: SE 4, SH per SE 1, CU per SH 16, active_cu_number 64` `[38724.906590] amdgpu 0000:09:00.0: amdgpu: ring gfx uses VM inv eng 0 on hub 0` `[38724.906592] amdgpu 0000:09:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0` `[38724.906593] amdgpu 0000:09:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0` `[38724.906594] amdgpu 0000:09:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0` `[38724.906596] amdgpu 0000:09:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0` `[38724.906597] amdgpu 0000:09:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0` `[38724.906598] amdgpu 0000:09:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0` `[38724.906599] amdgpu 0000:09:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0` `[38724.906600] amdgpu 0000:09:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0` `[38724.906601] amdgpu 0000:09:00.0: amdgpu: ring kiq_0.2.1.0 uses VM inv eng 11 on hub 0` `[38724.906603] amdgpu 0000:09:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 8` `[38724.906604] amdgpu 0000:09:00.0: amdgpu: ring sdma0 shares VM invalidation engine 0 with ring page0 on hub 8` `[38724.906606] amdgpu 0000:09:00.0: amdgpu: ring page0 uses VM inv eng 1 on hub 8` `[38724.906607] amdgpu 0000:09:00.0: amdgpu: ring sdma1 uses VM inv eng 4 on hub 8` `[38724.906608] amdgpu 0000:09:00.0: amdgpu: ring sdma1 shares VM invalidation engine 4 with ring page1 on hub 8` `[38724.906609] amdgpu 0000:09:00.0: amdgpu: ring page1 uses VM inv eng 5 on hub 8` `[38724.906610] amdgpu 0000:09:00.0: amdgpu: ring uvd_0 uses VM inv eng 6 on hub 8` `[38724.906611] amdgpu 0000:09:00.0: amdgpu: ring uvd_enc_0.0 uses VM inv eng 7 on hub 8` `[38724.906613] amdgpu 0000:09:00.0: amdgpu: ring uvd_enc_0.1 uses VM inv eng 8 on hub 8` `[38724.906614] amdgpu 0000:09:00.0: amdgpu: ring uvd_1 uses VM inv eng 9 on hub 8` `[38724.906615] amdgpu 0000:09:00.0: amdgpu: ring uvd_enc_1.0 uses VM inv eng 10 on hub 8` `[38724.906616] amdgpu 0000:09:00.0: amdgpu: ring uvd_enc_1.1 uses VM inv eng 11 on hub 8` `[38724.906617] amdgpu 0000:09:00.0: amdgpu: ring vce0 uses VM inv eng 12 on hub 8` `[38724.906618] amdgpu 0000:09:00.0: amdgpu: ring vce1 uses VM inv eng 13 on hub 8` `[38724.906619] amdgpu 0000:09:00.0: amdgpu: ring vce2 uses VM inv eng 14 on hub 8` `[38724.906919] amdgpu: Detected AMDGPU DF Counters. # of Counters = 8.` `[38724.906936] amdgpu: Detected AMDGPU 2 Perf Events.` `[38724.907224] amdgpu 0000:09:00.0: amdgpu: Runtime PM not available` `[38724.908792] amdgpu 0000:09:00.0: [drm] Registered 6 planes with drm panic` `[38724.908794] [drm] Initialized amdgpu 3.64.0 for 0000:09:00.0 on minor 1` `[38724.922406] fbcon: amdgpudrmfb (fb0) is primary device` `[38725.126667] amdgpu 0000:09:00.0: [drm] fb0: amdgpudrmfb frame buffer device` `[38725.137059] amdgpu 0000:12:00.0: enabling device (0000 -> 0003)` `[38725.137333] amdgpu 0000:12:00.0: amdgpu: initializing kernel modesetting (VEGA10 0x1002:0x6864 0x1002:0x0C00 0x05).` `[38725.137355] amdgpu 0000:12:00.0: amdgpu: Fatal error during GPU init` `[38725.137487] amdgpu 0000:12:00.0: probe with driver amdgpu failed with error -12` `[38725.137807] Modules linked in: amdgpu(+) amdxcp drm_panel_backlight_quirks gpu_sched drm_buddy drm_ttm_helper ttm drm_exec drm_suballoc_helper drm_display_helper cec rc_core i2c_algo_bit video xt_conntrack xt_MASQUERADE bridge stp llc xt_set ip_set nft_chain_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 xt_addrtype nft_compat nf_tables nfnetlink xfrm_user xfrm_algo snd_seq_dummy snd_hrtimer overlay zram 842_decompress 842_compress lz4hc_compress lz4_compress intel_rapl_msr intel_rapl_common intel_uncore_frequency intel_uncore_frequency_common skx_edac skx_edac_common nfit x86_pkg_temp_thermal intel_powerclamp snd_hda_codec_atihdmi coretemp snd_hda_codec_hdmi binfmt_misc snd_hda_intel snd_hda_codec dm_crypt apple_bce(C) nls_iso8859_1 snd_hda_core kvm_intel snd_intel_dspcfg snd_seq_midi snd_seq_midi_event snd_intel_sdw_acpi snd_rawmidi kvm snd_hwdep brcmfmac snd_seq snd_pcm brcmutil irqbypass snd_seq_device applesmc rapl cdc_acm snd_timer spi_nor intel_cstate cfg80211 snd mtd apple_mfi_fastcharge joydev mei_me` `[38725.138208] amdgpu_device_fini_sw+0x51a/0x700 [amdgpu]` `[38725.140174] amdgpu_driver_release_kms+0x16/0x40 [amdgpu]` `[38725.142190] ? __pfx_amdgpu_init+0x10/0x10 [amdgpu]` `[38725.144060] amdgpu_init+0x69/0xff0 [amdgpu]` `[38725.146198] amdgpu 0000:15:00.0: enabling device (0000 -> 0003)` `[38725.146409] amdgpu 0000:15:00.0: amdgpu: initializing kernel modesetting (VEGA10 0x1002:0x6864 0x1002:0x0C00 0x05).` `[38725.146427] amdgpu 0000:15:00.0: amdgpu: Fatal error during GPU init` `[38725.146515] amdgpu 0000:15:00.0: probe with driver amdgpu failed with error -12` It doesn't even work if it's the only card. It is never detected in Vulkan either. I have tried AMDVLK and RADV. I have tried it on x16 lanes and x8. I have two cards. Both have the same error. I did find one mention of there being a timing issue where something about the card doesn't activate right away so adding a timeout to the linux kernel would fix it. That also did nothing. Has anyone else experienced this?