Commits · a3edbec9a3e7c6668d5bdec2cbec2b31a1d5ec14 · gaoqiong / composable_kernel

30 Jan, 2021 1 commit
- debugging vector global store · a3edbec9
  Jing Zhang authored Jan 30, 2021
  
  a3edbec9
29 Jan, 2021 3 commits
- clean code · 631d9892
  Jing Zhang authored Jan 29, 2021
  
  631d9892
- fixed thread_buff r/w · 961556eb
  Jing Zhang authored Jan 29, 2021
  
  961556eb
- add initial in vector type; move buffer to gridwise gemm · 0486af77
  Jing Zhang authored Jan 29, 2021
  
  0486af77
21 Jan, 2021 3 commits
- clean code · 79137e1a
  Jing Zhang authored Jan 21, 2021
  
  79137e1a
- demo of removing array for A/B in xdlops · 63bad606
  Jing Zhang authored Jan 21, 2021
  
  63bad606
- fixed c initial · 494608ce
  Jing Zhang authored Jan 21, 2021
  
  494608ce
20 Jan, 2021 1 commit
- debug · 736a37ba
  Jing Zhang authored Jan 20, 2021
  
  736a37ba
11 Jan, 2021 1 commit
- debug matA vector load · 15232a0d
  Jing Zhang authored Jan 11, 2021
  
  15232a0d
05 Jan, 2021 1 commit
- clean · 7db237ba
  Jing Zhang authored Jan 05, 2021
  
  7db237ba
17 Dec, 2020 1 commit
- fixed buffer_load · 9d6938ff
  Jing Zhang authored Dec 17, 2020
  
  9d6938ff
16 Dec, 2020 5 commits
- debuging · 9f633f91
  Jing Zhang authored Dec 16, 2020
  
  9f633f91
- add original gridwise gemm · 888b7c78
  Jing Zhang authored Dec 16, 2020
  
  888b7c78
- fixed namming · 821ec5ae
  Jing Zhang authored Dec 16, 2020
  
  821ec5ae
- clean code · aaf3d81d
  Jing Zhang authored Dec 16, 2020
  
  aaf3d81d
- vector type output · 1d6022b1
  Jing Zhang authored Dec 16, 2020
  
  1d6022b1
15 Dec, 2020 4 commits
- clean code · 9a54fbd8
  Jing Zhang authored Dec 15, 2020
  
  9a54fbd8
- new c buffer type · be11c5dd
  Jing Zhang authored Dec 15, 2020
  
  be11c5dd
- simply vector type At · 649af8b2
  Jing Zhang authored Dec 15, 2020
  
  649af8b2
- tuning · 159559c1
  Jing Zhang authored Dec 15, 2020
  
  159559c1
14 Dec, 2020 2 commits
- add v4r4 xdlops · 7b002f23
  Jing Zhang authored Dec 14, 2020
  
  7b002f23
- adding xdlops · 87a75734
  Jing Zhang authored Dec 14, 2020
  
  87a75734
13 Dec, 2020 4 commits
- add vector store by 4 · 7972ab17
  Jing Zhang authored Dec 13, 2020
  
  7972ab17
- add vector store · 5494423f
  Jing Zhang authored Dec 13, 2020
  
  5494423f
- add other vector_sizes · acdafc67
  Jing Zhang authored Dec 13, 2020
  
  acdafc67
- add At into vector type · d5c856bc
  Jing Zhang authored Dec 13, 2020
  
  d5c856bc
12 Dec, 2020 5 commits
- clean code · 5d3b9b73
  Jing Zhang authored Dec 12, 2020
  
  5d3b9b73
- clean code · c85e0652
  Jing Zhang authored Dec 12, 2020
  
  c85e0652
- clean code · e64a79c5
  Jing Zhang authored Dec 12, 2020
  
  e64a79c5
- vector type demo for fp32 · 7bbcd0fe
  Jing Zhang authored Dec 12, 2020
  
  7bbcd0fe
- static load · bbdb77e8
  Jing Zhang authored Dec 12, 2020
  
  bbdb77e8
11 Dec, 2020 2 commits
- clean code · c1e24c09
  Jing Zhang authored Dec 11, 2020
  
  c1e24c09
- add thread_buff · 7abc0752
  Jing Zhang authored Dec 11, 2020
  
  7abc0752
03 Dec, 2020 2 commits
- revise comment · 82bf5de2
  Chao Liu authored Dec 03, 2020
  
  82bf5de2
- use StaticallyIndexedArray for buffer in threadwise copy, in order to get rid of alloca in IR · 760a234f
  Chao Liu authored Dec 03, 2020
  
  760a234f
30 Nov, 2020 1 commit
- fixing useless instruction issue · 70d06fa9
  Chao Liu authored Nov 30, 2020
  
  70d06fa9
28 Nov, 2020 1 commit
- use readfirstlane to force result into SGPR to reduce VGPR usage · 7733dd88
  Chao Liu authored Nov 28, 2020
  
  7733dd88
27 Nov, 2020 1 commit
- add blockwise copy that doesn't has thread buffer as member to avoid alloca... · 3b3cfae5
  Chao Liu authored Nov 27, 2020
```
add blockwise copy that doesn't has thread buffer as member to avoid alloca and therefore scratch mem
```
  3b3cfae5
26 Nov, 2020 2 commits
- update script · 54138dc8
  Chao Liu authored Nov 25, 2020
  
  54138dc8
- add dynamic forward v4r4, but see lots of scrach mem · be4e3133
  Chao Liu authored Nov 25, 2020
  
  be4e3133