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
- No tags were found...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
CUFFT - Fast Fourier Transform library• Algorithms based on Cooley-Tukey (n = 2a · 3b · 5c · 7d) <strong>and</strong> Bluestein• Simple interface similar to FFTW• 1D, 2D <strong>and</strong> 3D transforms of complex <strong>and</strong> real data• Row-major order (C-order) for 2D <strong>and</strong> 3D data• Single precision (SP) <strong>and</strong> Double precision (DP) transforms• In-place <strong>and</strong> out-of-place transforms• Batch execution for doing multiple transforms• Streamed asynchronous execution• Non normalized output: IFFT(FFT(A))=len(A)*A• Major constrains for better performance (single power, length multiplies)February 10, 2012 PRACE Winter School 20126