topic Re: migrating cusparseaxpby in Intel® oneAPI Math Kernel Library

migrating cusparseaxpby

sycl-developer — Thu, 07 Sep 2023 21:27:41 GMT

Is there an oneMKL function for https://docs.nvidia.com/cuda/cusparse/index.html#cusparseaxpby ? Thanks.

Re: migrating cusparseaxpby

JilaniS_Intel — Mon, 11 Sep 2023 11:57:42 GMT

Hi,

Thanks for posting in Intel Communities.

As per the requirement mentioned by you, it seems you are looking for a function which computes the sum of a sparse vector X and a dense vector Y. Could you please let us know if our understanding is correct?

Best Regards,

Jilani

Re: migrating cusparseaxpby

sycl-developer — Mon, 11 Sep 2023 15:56:34 GMT

Yes.

Re: migrating cusparseaxpby

Spencer_P_Intel — Mon, 11 Sep 2023 16:14:13 GMT

We currently do not have support for any API that includes a sparse vector including axpby, gather, scatter, dot product, etc. These may be considered for some future release but are not currently supported.

Spencer

Re: migrating cusparseaxpby

sycl-developer — Mon, 11 Sep 2023 16:16:52 GMT

Thank you for reading the cuSparse functions and considering them for future releases

Re: migrating cusparseaxpby

IntelSupport — Tue, 19 Sep 2023 11:42:05 GMT

Hi,

Thanks for helping us improve our products! We’ve submitted the feature request to the dev team Once it is included in an upcoming release, it would be documented in the release notes.

Best Regards,

Jilani Shaik