ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-07-02 13:17:36 +00:00
Code Issues Packages Projects Releases Wiki Activity
2,243 Commits 777 Branches 40 Tags
40ed20a30d043dc98ec426fc247a2aaa30d079bf
Commit Graph

509 Commits

Author SHA1 Message Date
Ding, Yi
331ccb8ca2 Merge remote-tracking branch 'origin/develop' into gfx950-mxfp4 2025-06-03 05:38:56 +00:00
Ding, Yi
0cd2e6e782 Fix OOB; add MB96 instances 2025-05-30 07:46:28 +00:00
Ding, Yi
69418725a6 Merge remote-tracking branch 'origin/moe_bs_fp8_no_asm' into gfx950-mxfp4 2025-05-30 03:15:47 +00:00
Ding, Yi
e4a40c7214 Add fp8 profiler instances 2025-05-29 08:19:31 +00:00
Bartłomiej Kocot
e7906dd644 Change relu to clamp for grouped conv fwd instances (#2249) 2025-05-29 00:51:25 +02:00
Adam Dickin
6df1c56ad6 Changes to allow MIOpen to build CK as part of its build. (#2247) 
* tweaks to the miopen specific build.  add way to skip clang-tidy checks and a way to skip some custom build targets MIOpen also has.

* move the tidy if statment

---------

Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
 2025-05-28 13:51:15 -07:00
BrianHarrisonAMD
e91be7d96a Add option to disable offload compress for CK builds (#2250) 
* Add option to disable offload compress for CK builds

* Remove gemm exe offload compress flag conditional
 2025-05-28 13:47:56 -07:00
Ding, Yi
aecac410d0 Merge remote-tracking branch 'origin/f8blk_scale_opt' into wip-f4-mergemoe-2 2025-05-28 11:15:22 +00:00
Ding, Yi
ad7fd89c1d Merge remote-tracking branch 'origin/feiw/mxfp4_moe_2Stages' into wip-f4 2025-05-28 09:28:26 +00:00
Ding, Yi
4f9bfb1566 Add more fp4 wp instances 2025-05-28 07:33:40 +00:00
Ding, Yi
857ef9f8c4 Merge preshuffle device 2025-05-28 07:02:28 +00:00
Ding, Yi
e2e0e0025e Profiler add f4 wp 2025-05-28 05:12:39 +00:00
Ding, Yi
b99c50a5d5 pad ascale 2025-05-28 03:35:33 +00:00
Ding, Yi
cf5b4c11a2 Pad shuffled a scale only 2025-05-28 02:37:14 +00:00
Bartłomiej Kocot
b1ed92b131 Revert "Remove not needed bwd wei merged groups instances (#2218)" (#2235) 
This reverts commit 4583aeffad.
 2025-05-26 23:26:04 +02:00
Bartłomiej Kocot
4583aeffad Remove not needed bwd wei merged groups instances (#2218) 
* Grouped conv bwd wei add two stage instances for larger filter and Merge Groups

* Fix

* fix

* Revert "Restore oddc instances (#2201)"

This reverts commit 6342f6b5e8.

* fix

---------

Co-authored-by: Rostyslav Geyyer <46627076+geyyer@users.noreply.github.com>
 2025-05-26 22:46:18 +02:00
Ding, Yi
91eb136937 Fix v1; use M padding 2025-05-26 10:32:26 +00:00
Andriy Roshchenko
fdfc9c6fd8 Merge remote-tracking branch 'origin/develop' into andriy/wip-f4 2025-05-23 23:02:43 +00:00
Andriy Roshchenko
f03da29b65 Merge branch origin/wip-f4 into andriy/wip-f4 2025-05-23 22:14:30 +00:00
Illia Silin
bc2551ac3b disable building device_mha_operations by default (#2225) 2025-05-22 14:03:04 -07:00
Adam Dickin
417a6b65b6 Add MIOPEN_REQ_LIBS_ONLY option for cmake to build only the libs MIOpen requires (#2224) 
* cut out anything we dont need for MIOpen to test

* refactor exclusion code to be more streamlined.
 2025-05-22 11:14:33 -07:00
Ding, Yi
ce50d4bd62 Fix fp4 ckProfiler 2025-05-22 09:39:22 +00:00
Andriy Roshchenko
e302ab8f0c Merge branch origin/develop into wip-fp4 2025-05-22 06:31:47 +00:00
Bartłomiej Kocot
ebc5a6ef87 Grouped conv bwd wei add for larger filter and Merge Groupes optimization (#2197) 
* Grouped conv bwd wei add two stage instances for larger filter and Merge Groups

* Fix

* fix

* Restore removed instances

---------

Co-authored-by: Rostyslav Geyyer <46627076+geyyer@users.noreply.github.com>
 2025-05-21 22:47:34 +02:00
OscarXu
fc9ef98e7b Add gemm2 64x128x128 asm. Fix BF16 ref. 2025-05-21 16:57:57 +08:00
Thomas Ning
1386924749 Add the instances for small sized GEMM in preshuffle and improve CMake Flag (#2212) 
* Add small instance, add the bug fix, & improve the example CMake

* clang format
 2025-05-20 15:05:08 -07:00
OscarXu
b146191da2 v2 tok_weight WIP 2025-05-20 16:34:07 +08:00
Ding, Yi
667a356cc3 Add mx fp4 pileline v1 instances 2025-05-20 03:28:24 +00:00
Andriy Roshchenko
57e0f5df29 MX GEMM - Expand MX MFMA Testing to BF8, FP6, and BF6 Data Types (#2199) 
* Unify test interface for different layouts.

* WIP: Introducing FP4/FP6/FP8 abstractions

* WIP: Introducing packed storage abstraction

* WIP: Introducing packed storage abstraction

* WIP: Improved support for FP6 data type

* Refactor packed storage for f6_t

* WIP: FP6 MFMA test

* Test if we correctly represent all FP6/FP4 numbers

* Additional output for failed FP4 test.

* More failing conversion tests

* Even more failing conversion tests

* Working FP6 MFMA tests

* Expand MX MFMA testing to BF8/6

* Update and verify MX MFMA test for packed types

* Fix fp4 and fp6 conversions on host

* Working MX MFMA tests for FP8/6/4

* Cleanup

* Add missing type

* Cleanup

* Final cleanup

* Restrict FP6/4 values output to CK_LOGGING=1

* Use CHAR_BIT instead of number 8

* Fix typo

* Remove FP6 and FP4 from the list of native types

---------

Co-authored-by: Rostyslav Geyyer <rosty.geyyer@amd.com>
 2025-05-19 16:52:51 -05:00
Ding, Yi
ec240f391a Fix mx f4 ckProfiler 2025-05-19 06:39:43 +00:00
mtgu0705
a4b5a374b9 Merge remote-tracking branch 'origin/wip-f4-pk' into mx_moe_f4_scale_shuffle 2025-05-17 09:49:24 -05:00
Bartłomiej Kocot
6342f6b5e8 Restore oddc instances (#2201) 2025-05-16 18:42:02 -07:00
Illia Silin
40668c9a99 Build and store CK library deb package for all targets daily. (#2196) 
* generate and store library package for all targets

* use ninja to build packages for all targets

* make sure to use ftime-trace when using ninja

* make sure build trace only runs on gfx9

* archive lib package and stash only library package
 2025-05-16 07:40:53 -07:00
OscarXu
ec8d00d58d mx_moe_fp4 ready for aiter with clang-format. 2025-05-16 04:09:26 -05:00
OscarXu
c5be9a501b v1 function pass. 2025-05-16 03:16:38 -05:00
Ding, Yi
70e0d94932 Add f4 profiler examples 2025-05-16 07:49:55 +00:00
Ding, Yi
dc30e7d025 Add f4 ckProfiler 2025-05-16 07:19:22 +00:00
Ding, Yi
c04d44b5f6 Merge remote-tracking branch 'origin/develop' into wip-f4 2025-05-16 07:11:26 +00:00
Bartłomiej Kocot
7c0e29cc0f Extend 64x64 with 4 waves instances for grouped conv bwd wei (#2187) 
* Extend 64x64 with 4 waves instnaces for grouped conv bwd wei

* Fix

* fix

* fix
 2025-05-15 16:21:34 +02:00
OscarXu
f70f778e27 v1 compile pass. Function not ready 2025-05-15 08:01:56 -05:00
OscarXu
68dbe558df compile error fix 2025-05-15 16:55:20 +08:00
OscarXu
17922821ec Add gemm1 v1 2025-05-15 16:11:43 +08:00
OscarXu
98606fad94 Merge remote-tracking branch 'origin/wjx/moe_v3_aiter' into moe_bs_stage1_dev 2025-05-14 20:48:42 -05:00
mtgu0705
102151ebcf temp save 2025-05-14 08:13:47 -05:00
aska-0096
74686bf008 tempsave. Almost all instances passed. 2025-05-14 10:15:59 +00:00
feifei14119
7be8730247 fix cpu ref 2025-05-14 10:22:18 +08:00
mtgu0705
6dfe24c53e updated 2025-05-13 04:15:53 -05:00
valarLip
eaebefb278 update moe v1 pipeline 2025-05-13 08:42:49 +00:00
OscarXu
9ead312164 Gemm1 GUFusion function pass. Perf WIP 2025-05-13 15:32:39 +08:00
mtgu0705
5ba86c210b updated and build passed 2025-05-13 14:49:37 +08:00