Daine McNiven

Hi @peterbell10, Thanks for bringing this up. I am also seeing the slowdown in gemm compared to gemv using our rocblas-bench tool. I'll add this to my list and get...

Just wanted to update this and let you know that we have changes in the works. The performance comparison for gemv over gemm with m == 1 || n ==...

Hi @FreddieWitherden Thanks for bringing up this issue, we are looking into it now and will get back to you as soon as we can with some answers. Thanks again,...

Hi @JorgeG94, thanks for opening this issue. hipBLAS is just a wrapper library for rocBLAS/cuBLAS backends. rocBLAS then uses the Tensile library for calls to gemm. Since you're looking for...

Hi @hkfuertes, Unfortunately your APU (gfx90c) is not currently supported in ROCm. You can take a look at the supported GPUs in the [ROCM 6.0.2 Documentation](https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.0.2/reference/system-requirements.html#supported-gpus). [Some people have had...

Hi @jinz2014, [`hipblasComputeType_t`](https://github.com/ROCm/hipBLAS/blob/docs/6.0.0/library/include/hipblas.h#L541) was added for the ROCm 6.0.0 release, and includes `HIPBLAS_COMPUTE_32F_FAST_TF32` which is the equivalent of `CUBLAS_COMPUTE_32F_FAST_TF32`. Note that rocBLAS does not have an equivalent computeType so will...

Hi again @jinz2014, I hope the HIPBLAS_V2 API was able to satisfy your needs regarding the compute types for gemmEx. Again, this will be the default behaviour of hipBLAS in...

HI @jinz2014, ILP64 interfaces for all L1 and L2 functions are supported in hipBLAS as of [ROCm 6.2](https://rocm.docs.amd.com/en/docs-6.2.0/). Most L1 functions were added in 6.1.0, and the remaining ILP64 interfaces...

Hi @jinz2014, thanks for raising this issue. This would have to first be done in rocBLAS, then hipBLAS would basically work automatically with the rocBLAS backend. It doesn't look like...

@jinz2014 mixed precision dot support for 32-bit input and 64-bit output/compute has been added in rocBLAS at https://github.com/ROCm/rocBLAS/commit/1ae11221aacfa776b5e7c9f1542441ff826a77c8. This uses the rocblas_dot_ex(...) api in rocBLAS, or the hipblasDotEx(...) api in...