Changes between Version 85 and Version 86 of DataParallel/BenchmarkStatus
- Timestamp:
- 12/05/10 17:02:00 (2 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
DataParallel/BenchmarkStatus
v85 v86 113 113 114 114 A: The core for the vectorised.seq version is equivalent to the vector version. We expect the backend has compiled it differently. Check this again with LLVM.[[br]] 115 B: The vectorised.par version runs faster than vectorised.seq because the latter has a duplicate loopcounter in the inner loop. We need a duplicate-loop-counter removal optimisation.115 B: The vectorised.par version runs faster than vectorised.seq because the latter has a duplicate counter in the inner loop. We need a duplicate-loop-counter removal optimisation. 116 116 117 117 '''Status''': fine[[br]]
