Commits · 0f2a2519cb6e7754247bd24e5808402d7827507f · Andreas Klöckner / loopy · GitLab

Mar 15, 2016
- For 1D arrays, allow undetermined shape in shape determination · a9951731
  Andreas Klöckner authored Mar 15, 2016
  
  a9951731
Mar 14, 2016
- Add Python sparse matvec example · 6d49fcad
  Andreas Klöckner authored Mar 14, 2016
  
  6d49fcad
Mar 03, 2016
- Move shared ispc tools to loopy.tools · 538e29e9
  Andreas Klöckner authored Mar 03, 2016
  
  538e29e9
Feb 26, 2016
- Don't hardcode the ispc binary location · 86d7b15b
  Andreas Klöckner authored Feb 25, 2016
  
  86d7b15b
- Remove unused harness codegen code · 727f0ddc
  Andreas Klöckner authored Feb 25, 2016
  
  727f0ddc
- ISPC run harness: Initialize arrays, bring performance in line with non-goldilocks ICC perf · 00e9cd3d
  Andreas Klöckner authored Feb 25, 2016
  
  00e9cd3d
- Fix botched timing in ISPC harness (vrrrroom?) · 872e78b8
  Andreas Klöckner authored Feb 25, 2016
  
  872e78b8
- More tweaking the ISPC stream harness · 943f16b2
  Andreas Klöckner authored Feb 25, 2016
  
  943f16b2
Feb 25, 2016
- ISPC stream harness seems to get ICC performance (fingers crossed :) · fb3a9f2c
  Andreas Klöckner authored Feb 25, 2016
  
  fb3a9f2c
- Tweak ispc test harness · 10d550be
  Andreas Klöckner authored Feb 24, 2016
  
  10d550be
Feb 19, 2016
- Add ispc stream harness · 4c99d2c8
  Andreas Klöckner authored Feb 18, 2016
  
  4c99d2c8
Jan 30, 2016
- Adjust hello-world example to make sense on ISPC · 842f346f
  Andreas Klöckner authored Jan 30, 2016
  
  842f346f
Jun 11, 2015
- Revamp how Fortran source processing is done, expose passes in Python transforms · d8be1eae
  Andreas Klöckner authored Jun 10, 2015
  
  d8be1eae
Feb 03, 2015
- Initial floopy absorption · 24a8c0f5
  Andreas Klöckner authored Feb 03, 2015
  
  24a8c0f5