[hipblaslt] Update gfx12 F8BS TN SAB Vector Gridbase yaml for ROCm 7.0#1057
Merged
vamovsik merged 7 commits intorelease/rocm-rel-7.0from Sep 22, 2025
Merged
[hipblaslt] Update gfx12 F8BS TN SAB Vector Gridbase yaml for ROCm 7.0#1057vamovsik merged 7 commits intorelease/rocm-rel-7.0from
vamovsik merged 7 commits intorelease/rocm-rel-7.0from
Conversation
hyoon1
force-pushed
the
users/hyoon1/navi4_sabv_7.0
branch
from
4cfd7fb to
cc7ff9b
Compare
hyoon1
force-pushed
the
users/hyoon1/navi4_sabv_7.0
branch
from
cc7ff9b to
1be2387
Compare
cmingch
approved these changes
Aug 5, 2025
hyoon1
force-pushed
the
users/hyoon1/navi4_sabv_7.0
branch
from
1be2387 to
0202d3f
Compare
slojosic-amd
added a commit
to ROCm/hipBLASLt
that referenced
this pull request
Aug 8, 2025
Contributor
|
@ROCm/hipblaslt-reviewers Can we get another review/approval? |
cmingch
approved these changes
Sep 22, 2025
Contributor
|
Please re-run tests and verify that they all pass. Right now the precheckin tests are failing on gfx 12, which this PR is changing. |
assistant-librarian bot
pushed a commit
to ROCm/hipBLASLt
that referenced
this pull request
Sep 22, 2025
[hipblaslt] Update gfx12 F8BS TN SAB Vector Gridbase yaml for ROCm 7.0 (#1057) Following up of ROCm/rocm-libraries#1054 #1941 These configs are required to support DeepSeek-R1-Distill FP8 models on gfx12. Previously, it was only applied to release-6.4 branch. To keep further support for these models, minimized kernels to add, and it generally performs similar or better than before. Conducted hipblaslt-test on both gfx1200 and gfx1201, and it passed all tests. [----------] Global test environment tear-down [==========] 39066 tests from 11 test suites ran. (364139 ms total) [ PASSED ] 39066 tests. hipBLASLt version: 100100 hipBLASLt git version: ROCm/rocm-libraries@879361da5cbf79495ff43db977eada9cfc7db360-dirty command line: hipblaslt-test Co-authored-by: Val Movsik <160653499+vamovsik@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Following up of #1054 ROCm/hipBLASLt#1941
These configs are required to support DeepSeek-R1-Distill FP8 models on gfx12.
Previously, it was only applied to release-6.4 branch. To keep further support for these models, minimized kernels to add, and it generally performs similar or better than before.
Conducted hipblaslt-test on both gfx1200 and gfx1201, and it passed all tests.
[----------] Global test environment tear-down
[==========] 39066 tests from 11 test suites ran. (364139 ms total)
[ PASSED ] 39066 tests.
hipBLASLt version: 100100
hipBLASLt git version: 879361da5cbf79495ff43db977eada9cfc7db360-dirty
command line: hipblaslt-test