Skip to content
GitLab
  1. 20 Aug, 2016 1 commit
  3. 11 Aug, 2016 1 commit
  5. 01 Aug, 2016 2 commits
  7. 22 Jul, 2016 1 commit
  9. 15 Jul, 2016 4 commits
  11. 14 Jul, 2016 1 commit
  13. 04 Jul, 2016 2 commits
  15. 30 Jun, 2016 1 commit
  17. 29 Jun, 2016 1 commit
  19. 14 Jun, 2016 1 commit
  21. 14 Apr, 2016 1 commit
  23. 07 Apr, 2016 1 commit
  25. 02 Apr, 2016 1 commit
  27. 07 Mar, 2016 3 commits
  29. 05 Mar, 2016 1 commit
  31. 01 Mar, 2016 1 commit
  33. 29 Feb, 2016 1 commit
  35. 26 Feb, 2016 1 commit
  37. 24 Feb, 2016 2 commits
  39. 05 Feb, 2016 2 commits
  41. 21 Jan, 2016 1 commit
    • Angus Lepper's avatar
      NUMA-aware affinity for workers · 61b0165e
      Angus Lepper authored 
      * This is not vectorisation-specific.
* There may be better trade-offs between HT and NUMA.
* Maybe print a warning when we require multiple NUMA nodes.
* Must also detect when Hyper-Threading is not present.
* Probably better as a configure flag, rather than conditional only upon
  the availability of libnuma.
* ~15-40% performance improvement on COSMA.
      61b0165e
  43. 11 Jan, 2016 1 commit
  45. 07 Jan, 2016 2 commits
  47. 27 Nov, 2015 1 commit
  49. 22 Sep, 2015 1 commit
  51. 17 Sep, 2015 3 commits
  53. 16 Sep, 2015 2 commits