topic Re: Re:Adding vectors using oneMKL in Intel® oneAPI Math Kernel Library

Adding vectors using oneMKL

Nikhil_T — Wed, 09 Jun 2021 12:54:15 GMT

I want to add 2 vectors using the following command:

oneapi::mkl::vm::add ( general parameters )

The function signature says that it adds vec_a and vec_b and stores the result in vec_y.

However, in my case, I want to perform vec_a = vec_a + vec_b, multiple times. I want to avoid the creation of a new vec_y and then transfer it to vec_a.

Is it possible to call the function with the following signature:

(vec_a , vec_b and vec_y are std::vectors)

oneapi::mkl::vm::add(q, vec_size, vec_a.data(), vec_b.data(),vec_a.data(), {})

instead of

oneapi::mkl::vm::add(q, vec_size, vec_a.data(), vec_b.data(),vec_y.data(), {})

Also, when trying to add those 2 vectors, I am getting the error as :

E0304 no instance of overloaded function "oneapi::mkl::vm::add" matches the argument list

Re:Adding vectors using oneMKL

MRajesh_intel — Wed, 09 Jun 2021 13:47:43 GMT

Hi,

Thanks for posting your query.

>>oneapi::mkl::vm::add(q, vec_size, vec_a.data(), vec_b.data(),vec_a.data(), {})

We tried this scenario using CPU as the device and the results were correct.

>>Also, when trying to add those 2 vectors, I am getting the error :

Please refer to the examples provided in /intel/oneapi/mkl/latest/examples/dpcpp/vml. In the directory you can refer to vadd.cpp or vadd_usm.cpp for reference.

Please let us know if you have any issues.

Regards

Rajesh.

Re:Adding vectors using oneMKL

MRajesh_intel — Tue, 15 Jun 2021 08:31:18 GMT

Hi,

Can you please update us on whether your issue has been resolved or not?

Regards

Rajesh.

Re: Re:Adding vectors using oneMKL

Nikhil_T — Tue, 15 Jun 2021 08:40:50 GMT

@MRajesh_intel Thank you. Yes the issue has been resolved now. It was the older version of oneapi that was getting stuck.

Re:Adding vectors using oneMKL

MRajesh_intel — Tue, 15 Jun 2021 12:20:28 GMT

Hi,

Thanks for the confirmation!

As this issue has been resolved, we will no longer respond to this thread. If you require any additional assistance from Intel, please start a new thread. Any further interaction in this thread will be considered community only.

Have a Good day.

Regards

Rajesh