AMD has great hardware, but their software is a different story. It’s poorly documented, unstable, and doesn’t deliver good performance for end users.

I’ve been working with the AMD MI300X for a few weeks, trying to get matrix multiplication running with tools like CK, Triton, or hipBLAS. However, the performance is only about 50% of the theoretical peak (FP16: 650 TFLOPS/s vs. 1300 TFLOPS/s in the whitepaper). Note that this is with matrices initialized to zero. When using random floats, performance drops by 20%—this is confirmed in AMD’s documentation.

Meanwhile, the H100, MI300X’s competitor, has a theoretical FP16 performance of 1000 TFLOPS, and I can achieve 800-900 TFLOPS with matrix multiplication using CUTLASS and random floats initialization.

AMD needs to improve their software quickly if they want to catch up with NVIDIA.

Doesn't seem to be released yet. https://github.com/ROCm/llvm-project does not have a 6.3 tag.

Same for https://github.com/ROCm/rocm_smi_lib/releases

Anyone know how to find the list of AMD GPU/Accelerator hardware that ROCm 6.3 supports? Usually AMD drops an old line or two every time they update ROCm.

https://rocm.docs.amd.com/projects/radeon/en/latest/docs/com...

When looking at the latest support matrix it basically only supports these bleeding edge cards, "AMD Radeon RX 7900 XTX, AMD Radeon RX 7900 XT, AMD Radeon RX 7900 GRE, AMD Radeon PRO W7900, AMD Radeon PRO W7900DS, AMD Radeon PRO W7800".

Surely I'm misinterpreting this and that can't be all the cards they support with latest ROCm. Does anyone know a more complete list?