Skip to content
Snippets Groups Projects
  1. Jul 29, 2013
  2. Jul 07, 2013
  3. Jul 06, 2013
  4. Jun 10, 2013
  5. Mar 20, 2013
  6. Oct 24, 2012
    • dhj's avatar
      Updated benchmark to use equivalent work loads for cpu and gpu. · fa833504
      dhj authored
      * Numpy does element-wise operations by default.  Updated the cpu
        operation to use pure numpy.
      
      * Eliminated the loop which is not necessary to demonstrate
        parallelism on array operations.
      
      * Made the number of workers explicit rather than gpu chosen,
        through local_size variable passed to kernel execution.
      
      * Increased to ~8 million data points to more clearly demonstrate
        the difference between cpu and gpu based computations.
      fa833504
  7. Oct 06, 2012
  8. Aug 11, 2012
  9. Jul 30, 2012
  10. Jul 25, 2012
  11. Jun 18, 2012
  12. Dec 07, 2011
  13. Aug 22, 2011
  14. Aug 08, 2011
  15. Jul 04, 2011
  16. May 07, 2011
  17. Apr 29, 2011
  18. Apr 17, 2011
  19. Feb 25, 2011
  20. Feb 18, 2011
  21. Feb 17, 2011
  22. Jan 19, 2011
  23. Jan 18, 2011
  24. Jan 01, 2011
  25. Dec 15, 2010
  26. Oct 31, 2010
  27. Oct 20, 2010
  28. Sep 17, 2010
  29. Aug 06, 2010
  30. Jul 20, 2010
  31. Jul 16, 2010
  32. Jul 09, 2010
Loading