ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-19 12:30:16 +00:00
Code Issues Packages Projects Releases Wiki Activity
211 Commits 746 Branches 38 Tags
45e1ad4dea7d884ed4497f3bcf66e9ca07c1cd20
Commit Graph

42 Commits

Author SHA1 Message Date
Chao Liu
45e1ad4dea adding ConstantMergedTensorDescriptor, refactering ConstantTensorDescriptor, Sequence 
[ROCm/composable_kernel commit: acd7082fe1]
 2019-05-21 16:17:58 -05:00
Chao Liu
dec8c3ebdd adding implicit gemm v3 
[ROCm/composable_kernel commit: 33b5a8556b]
 2019-05-16 22:23:18 -05:00
Chao Liu
ac7741cc7c adding implicit gemm v3 
[ROCm/composable_kernel commit: b7d052459d]
 2019-05-15 09:58:17 -05:00
Chao Liu
aeeefc1de3 refactored 
[ROCm/composable_kernel commit: 4957d5a399]
 2019-05-02 14:49:20 -05:00
Chao Liu
8b7eafe959 implicit gemm v1r2: only load 1d filter 
[ROCm/composable_kernel commit: 00899f191b]
 2019-04-13 11:19:17 -05:00
Chao Liu
2fc34a9169 tuned implicit gemm v1 for 3x3 on AMD to 82%. Fixed a bug in 4d tensor blockwise copy. 
[ROCm/composable_kernel commit: 96ee9571e2]
 2019-04-10 18:10:18 -05:00
Chao Liu
35140bbc9e enabled ds_read_b128 and ds_write_b128 on hip c++ 
[ROCm/composable_kernel commit: e624df922d]
 2019-04-09 19:05:44 -05:00
Chao Liu
4fba215ede refactor 
[ROCm/composable_kernel commit: 1bd880a674]
 2019-04-09 16:13:14 -05:00
Chao Liu
20df62cc31 clean up 
[ROCm/composable_kernel commit: 5245a0162b]
 2019-04-06 16:27:07 -05:00
Chao Liu
a3f850c5e6 debugging 
[ROCm/composable_kernel commit: f6cb5b846d]
 2019-04-06 15:10:40 -05:00
Chao Liu
008c584abb Merge branch 'master' into inline_asm_v2 
[ROCm/composable_kernel commit: 605afd0fb6]
 2019-04-04 18:40:23 -05:00
Jing Zhang
9c54da52e8 clean code 
[ROCm/composable_kernel commit: 62c4d5dff3]
 2019-04-04 14:52:42 -05:00
Jing Zhang
43283f6452 unroll k 
[ROCm/composable_kernel commit: 313f3c07d2]
 2019-04-04 11:43:37 -05:00
Chao Liu
6bf787f524 enable 128x128 block gemm 
[ROCm/composable_kernel commit: 05d7a0875c]
 2019-04-03 19:04:17 -05:00
Jing Zhang
7bea559923 add 
[ROCm/composable_kernel commit: 6a3f3f951d]
 2019-04-03 17:08:14 -05:00
Jing Zhang
d0996d2840 increase depth of pipeline 
[ROCm/composable_kernel commit: 0de4286a4f]
 2019-04-03 15:19:40 -05:00
Jing Zhang
73661b97c0 inline asm 
[ROCm/composable_kernel commit: 0d6aa311e9]
 2019-04-03 14:34:41 -05:00
Jing Zhang
1c8570bdf5 refactor inline asm 
[ROCm/composable_kernel commit: 753b98b59a]
 2019-04-03 13:23:56 -05:00
Chao Liu
e423954e6e puting gridwise convolution into its own class 
[ROCm/composable_kernel commit: 6290e0b080]
 2019-04-02 20:18:01 -05:00
Chao Liu
fdb7d41661 cleaning up dead code 
[ROCm/composable_kernel commit: bdbc0eaad1]
 2019-04-02 17:58:44 -05:00
Jing Zhang
6f8247cf59 4x4 
[ROCm/composable_kernel commit: 114fdb58af]
 2019-04-01 17:02:02 -05:00
Chao Liu
7cbd63b2d0 refactor 
[ROCm/composable_kernel commit: e43d7bc63c]
 2019-04-01 15:17:22 -05:00
Jing Zhang
ba38e7b9ea merged ds_read and gemm, but register allocation is mess 
[ROCm/composable_kernel commit: d058d16407]
 2019-03-29 14:37:36 -05:00
Jing Zhang
1444c6c852 in press 
[ROCm/composable_kernel commit: d700ce86d3]
 2019-03-29 14:27:49 -05:00
Jing Zhang
b12b2bfbe3 inline gemm 8x8 
[ROCm/composable_kernel commit: 1d9b970186]
 2019-03-29 11:09:03 -05:00
Jing Zhang
ed14d29657 in progress 
[ROCm/composable_kernel commit: 57a8ccf3ba]
 2019-03-28 21:53:24 -05:00
Jing Zhang
2c1f45f86d in progress 
[ROCm/composable_kernel commit: 66d5e5b344]
 2019-03-28 21:08:29 -05:00
Jing Zhang
b7b9817d57 fixed conflict 
[ROCm/composable_kernel commit: f7498d66f9]
 2019-03-28 20:01:53 -05:00
Jing Zhang
534a4a0916 inline 
[ROCm/composable_kernel commit: 5fbf4f33d3]
 2019-03-28 20:00:31 -05:00
Chao Liu
2f8828e8d5 Jing's ds_read inline asm 
[ROCm/composable_kernel commit: d6d9a8e4ce]
 2019-03-28 19:46:29 -05:00
Jing Zhang
866156db04 fused functions 
[ROCm/composable_kernel commit: 2058bec8cf]
 2019-03-28 18:47:32 -05:00
Chao Liu
cd883e7581 experimenting 
[ROCm/composable_kernel commit: 766b0a9eaf]
 2019-03-24 12:09:57 -05:00
Chao Liu
6f1651f8a7 experimenting 
[ROCm/composable_kernel commit: f35c64eb78]
 2019-03-23 19:35:31 -05:00
Chao Liu
68514d0739 adding inline asm for 16x4 gemm 
[ROCm/composable_kernel commit: 22114959da]
 2019-03-22 19:42:38 -05:00
Chao Liu
971442aab2 experiment with hip compiler 
[ROCm/composable_kernel commit: 68ae0731f1]
 2019-03-22 19:40:01 -05:00
Chao Liu
2c3bd06d25 Merge branch 'direct_fp16' 
[ROCm/composable_kernel commit: fdaaaa500c]
 2019-03-22 16:46:41 -05:00
Chao Liu
3a29ca8414 adding assembly 
[ROCm/composable_kernel commit: 18a81e356e]
 2019-03-22 16:33:04 -05:00
Chao Liu
1f925812b2 hip build 
[ROCm/composable_kernel commit: 8c923db423]
 2019-03-22 14:22:58 -05:00
Chao Liu
7bfe330532 update hip build 
[ROCm/composable_kernel commit: 2c9b8c2432]
 2019-03-12 17:20:11 -05:00
Chao Liu
1c962a13ee device_implicit_gemm_convolution_1_chwn_csrk_khwn: use tensor copy (instead of pointwise) for writing output, 3x3 increased from 78% to 84%, 5x5 from 80% to 84% 
[ROCm/composable_kernel commit: a65ef90308]
 2019-02-19 11:47:46 -06:00
Chao Liu
13388cef47 add anther verision of batch gemm 
[ROCm/composable_kernel commit: 1cb9885058]
 2019-02-17 01:50:57 -06:00
Chao Liu
c0baa18a3f change file extension to hip.hpp and hip.cpp 
[ROCm/composable_kernel commit: b2888adfbe]
 2019-02-15 02:13:21 -06:00