How can I call the functions in the cuSPARSE library in a device function?

Question

I want to call the sparse matrix multiplication function in cuSPARSE library inside the kernel instead of directly calling it at the host side. I write a __device__ function to implement it.My CUDA is 11.3 and My hardware is V100.My code all follows NVIDIA CUDALibrarySamples: spmm_csr

But it fails with:

error: calling a __host__ function("cusparseSpMM") from a __device__ function("spmm_csr") is not allowed

How can I call it in __device__ function?

Or there are others ways to implement sparse matrix multiplication inside kernel?

talonmies · Accepted Answer · 2022-12-05 07:50:50Z

1

I want to call the sparse matrix multiplication function in cuSPARSE library inside the kernel instead of directly calling it at the host side

That is not possible. cuSPARSE is a host only library.

answered Dec 5, 2022 at 7:50

community wiki

talonmies

Sign up to request clarification or add additional context in comments.

3 Comments

ben286 Over a year ago

Thanks！Do you know whether the cuBLAS library is a host only library？

talonmies Over a year ago

Yes, cuBLAS is now also host only. There was a period where there was also a device side version of cuBLAS, but that was discontinued and removed from the CUDA toolkit several major release cycles ago

paleonix Over a year ago

@ben286 Some linear algebra features (mostly GEMM) for device code are available through CUTLASS.

Collectives™ on Stack Overflow

How can I call the functions in the cuSPARSE library in a device function?

1 Answer 1

3 Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Related