Andrew Chapman

Hello @FreddieWitherden, rocBLAS functions are not safe to use with HIP Graph functions. We will work towards making them Graph safe in future releases of rocBLAS.

AFAIK it requires creating a pool of memory associated with the graph. Nodes in the graph asynchronously allocate from the pool, after the allocation is successful kernels are launched asynchronously,...

There are comments above comparing gemmt to syr2k. I think it is more similar to syrkx provided by cuBLAS and rocBLAS. See the links below: - https://docs.nvidia.com/cuda/cublas/#cublas-t-syrkx - https://docs.amd.com/bundle/rocBLAS-User-Guide---rocBLAS-documentation/page/API_Reference_Guide_80.html#rocblas-xsyrkx-batched-strided-batched syrkx...

Hello @torrance, cublasCherkEx() supports `CUDA_C_8I` datatype for matrix A. This is a complex number with two 8 bit signed integers. I have some questions about this datatype: - Do you...

Hi @torrance , Thank you for the context and the use case. I was guessing this is related to radio astronomy and the installations you have in Western Australia.

Hello @jinz2014. We do not have a compelling enough use case to add axpby. We could reconsider adding axpby and reopen this issue if we knew of a use case....