yes lmk how i can help. at the minimum i can get you hw and help with PRs etc. firstname at amd.com to reach me.

Im running ROCm ok on my 9070XT. You can build it from source today if you have a card.

rocminfo:

**** Agent 2 **** Name: gfx1201 Uuid: GPU-cea119534ea1127a Marketing Name: AMD Radeon Graphics Vendor Name: AMD Feature: KERNEL_DISPATCH Profile: BASE_PROFILE Float Round Mode: NEAR Max Queue Number: 128(0x80) Queue Min Size: 64(0x40) Queue Max Size: 131072(0x20000)

[32.624s](rocm-venv) a@Shark:~/github/TheRock$ ./build/dist/rocm/bin/rocm-smi

======================================== ROCm System Management Interface ======================================== ================================================== Concise Info ================================================== Device Node IDs Temp Power Partitions SCLK MCLK Fan Perf PwrCap VRAM% GPU% (DID, GUID) (Edge) (Avg) (Mem, Compute, ID) ================================================================================================================== 0 2 0x73a5, 59113 N/A N/A N/A, N/A, 0 N/A N/A 0% unknown N/A 0% 0% 1 1 0x7550, 24524 36.0°C 2.0W N/A, N/A, 0 0Mhz 96Mhz 0% auto 245.0W 4% 0% ================================================================================================================== ============================================== End of ROCm SMI Log ===============================================

ROCm on Radeon should work too and the poll above was to seek feedback on what to cards to support next.

I will provide this feedback to the docs team to clean up. I found it hard when i was making that Poll :D but I looked harder instead of trying to fix the docs. So thank you for the feedback.

Thanks for the effort! Documentation is so easy to neglect especially when there's also a ton of other stuff to do, but it's also so important for anything intended for technical use. Especially when things change over time or from one version to another.

Is this the repo you are referring to https://github.com/amd/go_amd_smi ? Would having a prebuilt version there help you ?

“ * NOTE: * The GO SMI binding depends on the following libraries:

- E-SMI inband library ("https://github.com/amd/esmi_ib_library") - ROCm SMI library("https://github.com/ROCm/rocm_smi_lib") - AMDSMI library("https://github.com/ROCm/amdsmi") - goamdsmi_shim library ("https://github.com/amd/goamdsmi/goamdsmi_shim")”

First of all this link is dead: https://github.com/amd/goamdsmi/goamdsmi_shim

Second: these dependencies should all be packaged into deb/rpm

Third: there should be a goamdsmi package which has a proper dependency tree. I should be able to do ‘apt-get install goamdsmi’ and it should install everything I need. This is how it works with go-nvml.

yes. We are behind on software support for all consumer cards and would love to support all cards. But are looking for guidance / feedback so we can prioritize.

This line sparks no confidence:

> No guarantees of future support but we will try hard to add support.

AMD reps told me exactly the same thing years ago about how they'd love to support all cards, when RDNA2 had just launched. Fast forward, only W6800 is properly supported from that gen. The last time I tried, it had tons of kernel bugs that caused hard freezes outside most basic cases.

You need to come out and say that you will support all cards, no ifs or buts, by a hard deadline.

I have quad w7900s under my desk that work well for workloads on my desktop that translate well to MI300x. There are some perf gaps with FAv2, and FP8 but otherwise I get a seamless experience. lmk if you have a pointer to any github issues for me to track down to make your experience better.

Do you not have (remote) access to a MI300X?

We do care about software and acknowledge the gaps and will work hard to make it better. Please let me know any specific issues that are an issue for you and Im happy to push for it to get resolved or come back with why it isn't.

PTX does provide a low level machine abstraction. However you still target some version of hardware ( https://arnon.dk/matching-sm-architectures-arch-and-gencode-... ). However a lot of software effort has gone into it to make it look and work seamlessly.

Though AMD doesn't have the same "virtual ISA" as PTX right now there are increasing levels of such abstraction available in compiled flows with MLIR / Linalg etc. Those are higher level and can be compiled / jitted in realtime to obviate the need for a low level virtual ISA.

hey thats me. Happy to help answer anything here and look forward to your constructive feedback to make AMD software better. We got work to do and look forward to it.

Ok, why does running koboldcpp with a "BLAS Batch Size" of 512 via Vulkan on an RX570 crash my entire computer? You know, to the point where I manually have to turn it on again.

I personally couldn't think of a better reason to never buy AMD GPUs ever again by the way.

I have experience running 130,000 RX470/570/480/580... if you're doing heavy workloads, those things full machine crash if you breathe on them wrong. That said, when they do run, they run extremely well.

There is 1000 reasons why your one GPU could have crashed, what does it say in the logs before it crashed?

FWIW crashing a machine because of your GPU driver is pretty insane. Why would you want to use a stack like this?