Commits · main · Stepan Nassyr / uarch_benchmarks

09 Sep, 2021 1 commit
- loopbegin/end now take a label name · b2f970b6
  Stepan Nassyr authored Sep 09, 2021
  
  b2f970b6
03 Sep, 2021 1 commit
- avx fma operand order · a9b83716
  Stepan Nassyr authored Sep 03, 2021
  
  a9b83716
31 Jul, 2021 1 commit
- indepfma: dep. breaking (commented out) · cc28708d
  Stepan Nassyr authored Jul 31, 2021
  
  cc28708d
30 Jul, 2021 12 commits
- AVX512 has 32 vregs, not 16 · d9a2abcb
  Stepan Nassyr authored Jul 30, 2021
  
  d9a2abcb
- neon: remove some unused code · 40828aab
  Stepan Nassyr authored Jul 30, 2021
  
  40828aab
- clarify what $GETSIMDSIZE gets replaced with · 5df9c75c
  Stepan Nassyr authored Jul 30, 2021
  
  5df9c75c
- Update GEMM.md · d2e2e7b0
  Stepan Nassyr authored Jul 30, 2021
  
  d2e2e7b0
- GEMM.md: add deps and supported platforms · 0b431eca
  Stepan Nassyr authored Jul 30, 2021
  
  0b431eca
- GEMM.md: fix title · 3dfa19b4
  Stepan Nassyr authored Jul 30, 2021
  
  3dfa19b4
- Fix Math in GEMM.md · d0111f6b
  Stepan Nassyr authored Jul 30, 2021
  
  d0111f6b
- Add ; to what's supposed to be c/c++ code · d2684a39
  Stepan Nassyr authored Jul 30, 2021
  
  d2684a39
- Merge branch 'wip_gemmerator' of... · dd461779
  Stepan Nassyr authored Jul 30, 2021
```
Merge branch 'wip_gemmerator' of gitlab.jsc.fz-juelich.de:nassyr1/fma_indep_bench into wip_gemmerator
```
  dd461779
- Update notebooks · ceaf6e66
  Stepan Nassyr authored Jul 30, 2021
  
  ceaf6e66
- Update GEMM.md · 8206ff47
  Stepan Nassyr authored Jul 30, 2021
  
  8206ff47
- Update README/documentation, README.md as an overview, specific benchmarks in their own files · 7b9116d5
  Stepan Nassyr authored Jul 30, 2021
  
  7b9116d5
29 Jul, 2021 1 commit
- document nanogemm function · 4817ab53
  Stepan Nassyr authored Jul 29, 2021
  
  4817ab53
28 Jul, 2021 5 commits
- avx: deduplicate code with inheritance · 8e002030
  Stepan Nassyr authored Jul 28, 2021
  
  8e002030
- AVX512 fixes · 52953c23
  Stepan Nassyr authored Jul 28, 2021
  
  52953c23
- Improve automatic unroll calculation · ff68c83f
  Stepan Nassyr authored Jul 27, 2021
  
  ff68c83f
- Forgot to loop over vectors in mr when postloading a · 44a8aed9
  Stepan Nassyr authored Jul 28, 2021
  
  44a8aed9
- typos · fb0ccf35
  Stepan Nassyr authored Jul 28, 2021
  
  fb0ccf35
27 Jul, 2021 2 commits
- -A (avec_strat) switch for gemm shell script · b717036f
  Stepan Nassyr authored Jul 27, 2021
  
  b717036f
- Add avec_strat (postload+preload), crude x86/avx support, refactor fma... · e531d311
  Stepan Nassyr authored Jul 27, 2021
```
Add avec_strat (postload+preload), crude x86/avx support, refactor fma throughput bench (so it works)
```
  e531d311
23 Jul, 2021 10 commits
- Fix up neon indexed FMA + SGEMM calculations · 0dde762b
  Stepan Nassyr authored Jul 23, 2021
  
  0dde762b
- neon: fix load_vector_dist1_inc addressing mode · 3c8424e7
  Stepan Nassyr authored Jul 23, 2021
  
  3c8424e7
- use load_vector_dist1 correctly · d539e669
  Stepan Nassyr authored Jul 23, 2021
  
  d539e669
- implement load_vector_dist1 in neon and sve · b70e60c1
  Stepan Nassyr authored Jul 23, 2021
  
  b70e60c1
- Correct option for .by script · e726d493
  Stepan Nassyr authored Jul 23, 2021
  
  e726d493
- Correct option for .by script · 8439202e
  Stepan Nassyr authored Jul 23, 2021
  
  8439202e
- auto option for bvec_strat and support in shell script · 14a287bc
  Stepan Nassyr authored Jul 23, 2021
  
  14a287bc
- adding dist1_inc strategy (and support it with neon) · 02eb39dc
  Stepan Nassyr authored Jul 23, 2021
  
  02eb39dc
- Update .gitignore · 1abfa1a9
  Stepan Nassyr authored Jul 22, 2021
  
  1abfa1a9
- NEON works, but not well tested (might perform badly) · e3946580
  Stepan Nassyr authored Jul 22, 2021
  
  e3946580
22 Jul, 2021 4 commits
- gemmerator: WIP NEON · 5d2156e4
  Stepan Nassyr authored Jul 22, 2021
  
  5d2156e4
- Get SVE version working · 694e5a96
  Stepan Nassyr authored Jul 22, 2021
  
  694e5a96
- Initial gemm generator implementation + rework · f10b1927
  Stepan Nassyr authored Jul 22, 2021
  
  f10b1927
- Update gitignore · df9539bc
  Stepan Nassyr authored Jul 22, 2021
  
  df9539bc
14 Jul, 2021 3 commits
- Remove useless semicolons from python script · aff757a7
  Stepan Nassyr authored Jul 14, 2021
  
  aff757a7
- Change names of avx+fma options for correctness · 1a42c63f
  Stepan Nassyr authored Jul 14, 2021
  
  1a42c63f
- refactor generator in preparation for more benchmarks · 97a5ea1d
  Stepan Nassyr authored Jul 14, 2021
  
  97a5ea1d