Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix for Unsupported Input Shapes/Sizes in Stream-K GEMM - BF16/FP16 #1866

Draft
wants to merge 6 commits into
base: develop
Choose a base branch
from
Draft

Fix for Unsupported Input Shapes/Sizes in Stream-K GEMM - BF16/FP16 #1866

wants to merge 6 commits into from
+10,198 −677

Conversation

ozturkosu
Copy link
Contributor

@ozturkosu ozturkosu commented Feb 6, 2025 •
edited
Loading

Proposed changes

Problem

Current Implementation of BF16/FP16 Stream-K GEMM does not support some input configurations whose input sizes are not divisible by 4.

See meta_amd_ai_operations/results at main · ROCm/meta_amd_ai_operations for examples which are seen as "0"

Proposed Solution

Checklist

Please put an x into the boxes that apply. You can also fill these out after creating the PR. If you're not sure, please don't hesitate to ask.

  • I have added tests relevant to the introduced functionality, and the unit tests are passing locally
  • I have added the test to REGRESSION_TESTS list defined at the top of CMakeLists.txt in tests/CMakeLists.txt, IF the test takes more than 30 seconds to run.
  • I have added inline documentation which enables the maintainers with understanding the motivation
  • I have removed the stale documentation which is no longer relevant after this pull request
  • (If this change is user-facing) I have added release notes which provide the end users with a brief summary of the improvement from this pull request
  • I have run clang-format on all changed files
  • Any dependent changes have been merged

Discussion

illsilin and others added 3 commits February 4, 2025 12:35
@illsilin @aska-0096
@junliume @andriy-ca @geyyer @jefyang1
Merge from internal (#1857)
800cf89 
* enable batched_gemm_softmax_gemm_perm_wmma for gfx12

* disable instances with blocksize=256 in attention examples

* debuggging

* debug

* fixed lds_enabled

* debugging

* Fix and add limit to skiplds feature

* Enable skipLds feature and fix compilation bugs

* add ck_tile definitions for gfx12

* fix clang format and test/wmma_op

* updage instances cmake for gfx12

* disable the test_wmma_op on gfx12

* fix the builds for gfx950

* add gfx12 and gfx950 to default target list

* clean-up cmake file

* Initial introduction of OFP8 data types.

* Renamed FP8 and BF8 tests into FP8_FNUZ and BF8_FNUZ.

* Implementation of ConvertFP32Nearest in test_fp8_ocp.

* Remove dependence on possibly undeclared alias.

* Implement FP8OCP test for stochastic rounding mode.

* Implement FP8OCP tests for half_t type conversions.

* enable bf16 atomic add on gfx950

* Implement ConvertFP32Nearest test.

* Implement ConvertFP32Stochastic test.

* Implement ConvertFP16Nearest and ConvertFP16Stochastic tests.

* Refactoring. Move FP8 definitions into a separate header file.

* Enable easy switching between architectures.

* Fix compilation error for gfx942 architecture.

* Add fp4 type with constants

* only builf gfx950 branch for gfx950 target by default

* Enable OCP build of example_gemm_xdl_fp8.

* Fix formatting.

* fix the build logic for gfx950

* Improve GEMM example verbosity.

* Add constexpr where applicable.

* fix the logic of enabling XDL and WMMA instances

* Improve GEMM example verbosity.

* Enable build of example_gemm_xdl_fp8_bf8 test.

* Fix tests for gfx1101 architecture.

* Build DPP examples only on gfx103 and gfx11 architectures.

* Optionaly run either CPU or GPU verifications with GEMM examples.

* Extend GeneratorTensor_Sequential to produce values of prescribed data types.

* Add missing constructor.

* Add scale type and mxfp conversions

* Update conversions

* Add conversion tests

* Fix typo

* Improve infrastructure for OFP8 data type support.

* BUGFIX. Should not use FP8 as Compute/Accum data type.

* Add custom target for grouped_convnd_bwd_weight tests.

* Can build `tests` target on gfx950.

* Bugfixes on gfx1101 architecture.

* Fix dependencies.

* Add stochastic rounding tests

* Provide single point of truth for FP8 INF and NAN checks

* Prevent instantiation of operators that are not supported by FP8 data types

* Add FP8 type selection into client_axample CMakeLists.txt

* Prevent sccache server from shutting down during build

* Fix test success reporting logic

* Change default verification method to CPU.

GPU verification takes too much time to complete on the emulator.

* Add scale <-> float conversions

* Add scaled conversions with tests

* Add device conversions

* Make sure all tests and examples are built for gfx950

* Facilitate testing of FP8 data types on the emulator

* Introduce two new tensor generators

* Enable instances built for gfx94 to be built on gfx950

* Verify 35_splitk_gemm on floating point numbers.

splitk gemm appears to be losing precision VS reference implementation when FP numbers are involved.

* Format

* Verify 04_gemm_add_add_fastgelu on floating point numbers

* Verify 20_grouped_conv_bwd_weight on floating point numbers

* Verify 38_grouped_conv_bwd_data_multiple_d on floating point numbers

* Verify more tests on floating point data

* Fix data types and improve testing verbocity.

* Add fp4 vectors

* Add debug tests

* Upgrade to NPI 573 build docker.

* Skip on gemm_universal tests.

The tests take too long to complete on the emulator.
Need to see if it is possible to reduce the scope of the testing to just FP8 data types.

* Add new mfma instructions and examples

* Add preprocessor directives for gfx950 specific code

* Fix gfx1101 build

* Document test availability

* Re-enable fp8 gemms for gfx94/95

* Cherry-pick GEMM Universal tests for FP8 data types

* Cleanup

* Add vector types and tests

* Add check_err function

* Add tensor generators

* CK_USE_GFX94 has already been set on this branch

* Fix

* Address formatting issues and leftovers

* Make fail/pass logic consistent within 01_gemm folder

Removed multiple negations in fail/pass logic to propagate `true` as the success indicator.

* Fix GPU verification reporting logic.

* Update year in copyright notice.

* Cleanup

* Use `enum class` instead of `enum`

* Remove set_property for FP8 tests

* Add vector conversions

* Fix

* Fix linker errror

* Clean up

* Fix gfx950 conversions

* Clean up

* Fix more gfx950 conversions

* Fix even more gfx950 conversions

* Narrowing the scope of PR to OCP FP8 enablement only

* Add tests for OCP FP8 vector_type storage

* Fix client examples build

* Fix typo

* Update e8m0 casting

* Rename E8M0 type

* Update unpack method

* Cleanup merge artifacts

* Enable gemm kernel on all gfx9 architectures (#227)

* clean-up

* Implement `non_native_vector_base` with `ext_vector_type` array. (#232)

* Enable support of 1, 2, 4, and 8-byte custom types in CK.

* Fix pool tests for OCP FP8 data type

* Fix build

* Add ckProfiler gemm instances for new mfma instructions and fix ckProfiler build on MI350

* fix clang format

* Add new mfma instructions and examples

* Add preprocessor directives for gfx950 specific code

* Add ckProfiler gemm instances for new mfma instructions and fix ckProfiler build on MI350

* fix clang format

* Fix clang format for the newly merged files

* Use the existing example instances for fp16 bf16 and int8

* Remove comment on new mfma instructions in MfmaInstr

* Update include/ck/tensor_operation/gpu/grid/gridwise_batched_gemm_gemm_xdl_cshuffle_v1.hpp

Co-authored-by: Andriy Roshchenko <[email protected]>

* merge from public repo

* Fix ck build

* Fix ck build

* Use double for max_abs_in_val

* Move scaled_type_convert functions to a separate header (#251)

* re-enable building mha lib and gemm_universal_f8 instances for gfx950

* Update library/src/tensor_operation_instance/gpu/CMakeLists.txt

Co-authored-by: Andriy Roshchenko <[email protected]>

* fix typo for CK_USE_OCP_FP8

* fix typo for CK_USE_OCP_FP8

* Add FP6 and BF6 types (#261)

* Add a rounding flag

* Add FP6 and BF6

* Add tests

Co-authored-by: Andriy Roshchenko <[email protected]>

* Clean up

---------

Co-authored-by: Andriy Roshchenko <[email protected]>

* fix one more typo

* Refactor E8M0 scale implementation (#262)

* Refactor E8M0 scale implementation

* Add MXFP6 and MXBF6 conversion methods (#270)

* Add conversions

* Add tests

* Add docstrings

* Add scaled conversions

* Add fp6/bf6 tests

* Remove misleading fp4 test case

* Add docstrings

* Clean up

* Address comments

* Set stricter tolerances for RNE tests

* Add missing tests

* Add native conversions to float

* Revert "Add native conversions to float"

This reverts commit 09467111f73b753c8cc3d597533b187940353dab.

* Update copyright years

* replace the fp6 with bf6 convert calls in test_bf6

* fix test_bf6

* enable smfmac test

* [MX FP8] Add Scaled Type Convert Functions for OCP FP8/BF8 data types (#271)

* Move scaled_type_convert functions to a separate header

* Introduce MX data tests

* Build MX tests only on relevant architectures

* Refactor E8M0 scale implementation

* Fix `config.h` typo

* Cleanup deprecated symbols

* Refactor `amd_ck_fp8.hpp`

* `scaled_type_convert` for `f8_ocp_t`

* Implement test for MX FP8 scaled type convert

* Implement test for MX BF8 scaled type convert

* Scaled type convert for vectors of 2 FP8 elements

* Scaled type convert for vectors of 16 FP8 elements

* Implementation of scaled conversion from F32 to F8

* Add tests for scaled conversions from FP32 to FP8

* Add documentation to the test functions

* Implementation of scaled conversion from F32x2 to F8x2

* Implementation of scaled conversion from F32x16 to F8x16

* Implementation of scaled conversion from F32x32 to F8x32

* Implementation of scaled conversion from F8x32 to F32x32

* Verified on the emulator

* MX FP GEMM - Example Template (#277)

Temporarily uses `DeviceGemmMultiD_ABScale_Xdl_CShuffle_V3` kernel and 128x128 scaling matrices.
Must be modified to use MX-native GEMM kernell with 16 or 32 component vectors per scale.

Verified on the emulator.

* Add vector support

* Add tests

* Add missing type aliases

* Fix test naming

* only build mx example for gfx950

* disable CK_USE_AMD_MFMA_GFX950 by default

* fic build for multiple archs

* fix typo

* fix typo

* Update unpack signature

* Fix merge

* Add size checks in pack function

* Add a flag

* Add conversions

* Fix build logic

* Update pack/unpack methods

* Remove unneeded AsType accessors

* Add docstrings

* Add a flag to config file

* Test the functionality of V_MFMA_F32_16X16X128_F8F6F4 and  V_MFMA_F32_32X32X64_F8F6F4 instructions. (#293)

* Introduced MFMA tests

* Verified f8f6f4 MFMA Instructions

* Move flag logic to scaled_type_convert header

* Use pointers instead of array indices

* Fix a typo

* Update tests and pack functions

* Fix gemm gemm on gfx950

* Fix clang format

* restore the default gput target lists

* fix the jenkinsfile

* add missing ifdef

---------

Co-authored-by: Jing Zhang <[email protected]>
Co-authored-by: aska-0096 <[email protected]>
Co-authored-by: Jun Liu <[email protected]>
Co-authored-by: Andriy Roshchenko <[email protected]>
Co-authored-by: Rostyslav Geyyer <[email protected]>
Co-authored-by: Rostyslav Geyyer <[email protected]>
Co-authored-by: root <[email protected]>
Co-authored-by: Andriy Roshchenko <[email protected]>
Co-authored-by: jefyang1 <[email protected]>
Co-authored-by: jefyang1 <[email protected]>
@illsilin
restore cron trigger (#1863)
2bef550
@ozturkosu
bug fixed
a9a3a3e
@ozturkosu ozturkosu changed the title bug fixed Stream-k GEMM BF16/FP16 Non-Supported Input Shape/Size Fix Feb 6, 2025
@ozturkosu
Copy link
Contributor Author

I will review again and polish codes before converting PR !!! But the problem is resolved.

@ozturkosu ozturkosu self-assigned this Feb 6, 2025
@ozturkosu ozturkosu changed the title Stream-k GEMM BF16/FP16 Non-Supported Input Shape/Size Fix Fix for Unsupported Input Shapes/Sizes in Stream-K GEMM with BF16/FP16 Data Types Feb 6, 2025
@ozturkosu ozturkosu changed the title Fix for Unsupported Input Shapes/Sizes in Stream-K GEMM with BF16/FP16 Data Types Fix for Unsupported Input Shapes/Sizes in Stream-K GEMM - BF16/FP16 Feb 6, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@illsilin illsilin Awaiting requested review from illsilin illsilin will be requested when the pull request is marked ready for review illsilin is a code owner

@carlushuang carlushuang Awaiting requested review from carlushuang carlushuang will be requested when the pull request is marked ready for review carlushuang is a code owner

@qianfengz qianfengz Awaiting requested review from qianfengz qianfengz will be requested when the pull request is marked ready for review qianfengz is a code owner

@aosewski aosewski Awaiting requested review from aosewski aosewski will be requested when the pull request is marked ready for review aosewski is a code owner

@poyenc poyenc Awaiting requested review from poyenc poyenc will be requested when the pull request is marked ready for review poyenc is a code owner

@geyyer geyyer Awaiting requested review from geyyer geyyer will be requested when the pull request is marked ready for review geyyer is a code owner

@bartekxk bartekxk Awaiting requested review from bartekxk bartekxk will be requested when the pull request is marked ready for review bartekxk is a code owner

@andriy-ca andriy-ca Awaiting requested review from andriy-ca andriy-ca will be requested when the pull request is marked ready for review andriy-ca is a code owner

@afagaj afagaj Awaiting requested review from afagaj afagaj will be requested when the pull request is marked ready for review afagaj is a code owner

At least 1 approving review is required to merge this pull request.

Assignees

@ozturkosu ozturkosu

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@ozturkosu @illsilin @jakpiase