Is it possible to use the embedded Gpu with vml?

[ImpPhil]

If I'm using Nuget package "intelmkl.redist.win-x64" Version="2022.2.1.19754" from my C# benchmark test  PhilPJL/MklBenchmark (github.com) is it possible to have those operation (vdCos, vzMul etc) run on the GPU?

[ShanmukhS_Intel]

Hi Philip,

Thanks for posting on Intel Communities.

Could you please cross-verify if the mentioned GitHub link was correct or not? We tried going through that link. However, the requested page was not found. In addition, it would be helpful if you elaborate your on the details mentioned by you.

Best Regards,

Shanmukh.SS

[ImpPhil]

Oops, I created that repo as private.  It should be visible now.

I'm new to IntelMKL, but I read that various operations can be run on the GPU.  I have an 8700k so I was interested in comparing the speed of vector math library functions running on the GPU vs the CPU.  Is that possible?

[ShanmukhS_Intel]

Hi Philip,

Oops, I created that repo as private. It should be visible now.

>> Thanks for the update! We were able to access it now. We were able to clone and run the code shared via github.

Attached are the benchmark test results. Will update you on further details soon!

Best Regards,

Sanmukh.SS

[ShanmukhS_Intel]

Hi Philip,

We were able to set up the environment and run the code successfully. We were discussing your issue internally. We will get back to you with an update.

Best Regards,

Shanmukh.SS

[ShanmukhS_Intel]

Hi Philip,

As a similar question has been raised on the below thread as well with the same code base, could you please let us know if we could continue addressing this issue on the below thread and close this?

Reference URL:

https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/vzMul-performance/m-p/1441909#M34067

Best Regards,

Shanmukh.SS

[ShanmukhS_Intel · Accepted solution]

Hi Philip,

We are discussing your issue internally. We will get back to you with an update soon!

Regrets for the delay in response.

Best Regards,

Shanmukh.SS

[ShanmukhS_Intel]

Hi Philip,

For running vml functions on GPU, OpenMP offload works for C and Fortran applications and DPCPP interfaces only work for C/C++ applications.

Best Regards,

Shanmukh.SS

[ShanmukhS_Intel]

Hi Philip,

A gentle reminder:

We have heard back from you for a while. Could you please let us know if you need any other information or let us know if we could close this case at our end?

Best Regards,

Shanmukh.SS

[ShanmukhS_Intel]

Hi Philip,

Thanks for accepting our solution. If you need any additional information, please post a new question as this thread will no longer be monitored by Intel.

Best Regards,

Shanmukh.SS