OpenACC Use Case
From HP-SEE Wiki
(Difference between revisions)
(Created page with " == Matrix Matrix multiplication == We briefly discuss the development effort needed to implement a simple matrix matrix multiplication algorithm using OpenACC directives on GPU...") |
(→Matrix Matrix multiplication) |
||
Line 2: | Line 2: | ||
== Matrix Matrix multiplication == | == Matrix Matrix multiplication == | ||
- | We briefly discuss the development effort needed to implement a simple matrix matrix multiplication algorithm using OpenACC directives on GPU resources and showcase timing and performance results obtained via several development approaches (simple algorithm for CPU, BLAS, CUDA, CuBLAS and OpenACC). | + | We briefly discuss the development effort needed to implement a simple matrix matrix multiplication algorithm using OpenACC directives on GPU resources and showcase timing and performance results obtained via several development approaches (simple algorithm for CPU, usage of BLAS/LAPACK libraries, CUDA, CuBLAS and OpenACC). |
Revision as of 13:12, 15 July 2013
Matrix Matrix multiplication
We briefly discuss the development effort needed to implement a simple matrix matrix multiplication algorithm using OpenACC directives on GPU resources and showcase timing and performance results obtained via several development approaches (simple algorithm for CPU, usage of BLAS/LAPACK libraries, CUDA, CuBLAS and OpenACC).