13.07.2015 Views

CUDA Libraries and MPI+OpenMP+CUDA - Prace Training Portal

CUDA Libraries and MPI+OpenMP+CUDA - Prace Training Portal

CUDA Libraries and MPI+OpenMP+CUDA - Prace Training Portal

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>CUDA</strong> NEWD – Code Snippet (C)...#if defined __<strong>CUDA</strong>_3qecudaSafeCall( cudaMemcpy( dtmp_D, dtmp,sizeof( double ) * nspin_mag, cudaMemcpyHostToDevice ) );#endifblocksPerGrid = ( (nspin_mag) + 256 - 1) / 256;kernel_compute_deeq( qgm_D, deeq_D,aux_D, na, nspin_mag, ngm, nat, flag, ih, jh, nhm, omega,fact, qgm_na_D, dtmp_D );qecudaGetLastError("kernel launch failure");}}}} Poor parallelism (nspin_mag),<strong>CUDA</strong> kernel dropped!February 10, 2012 PRACE Winter School 201258

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!