Simplified NVFP4 quantize kernel for Torch API (#152)#152
Open
jwfromm wants to merge 1 commit intometa-pytorch:mainfrom
Open
Simplified NVFP4 quantize kernel for Torch API (#152)#152jwfromm wants to merge 1 commit intometa-pytorch:mainfrom
jwfromm wants to merge 1 commit intometa-pytorch:mainfrom
Conversation
jwfromm
force-pushed
the
export-D93169309
branch
from
7d6210e to
5f8c934
Compare
jwfromm
added a commit
to jwfromm/MSLK-1
that referenced
this pull request
Mar 10, 2026
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
jwfromm
added a commit
to jwfromm/MSLK-1
that referenced
this pull request
Mar 10, 2026
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
jwfromm
force-pushed
the
export-D93169309
branch
from
5f8c934 to
5a07c6e
Compare
jwfromm
added a commit
to jwfromm/MSLK-1
that referenced
this pull request
Mar 10, 2026
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
jwfromm
force-pushed
the
export-D93169309
branch
2 times, most recently
from
33966e0 to
a8a9ea5
Compare
jwfromm
added a commit
to jwfromm/MSLK-1
that referenced
this pull request
Mar 12, 2026
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
jwfromm
added a commit
to jwfromm/MSLK-1
that referenced
this pull request
Mar 12, 2026
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
jwfromm
force-pushed
the
export-D93169309
branch
from
a8a9ea5 to
ae1f096
Compare
Summary: This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests. Reviewed By: jiawenliu64 Differential Revision: D93169309
meta-codesync
bot
changed the title
jwfromm
force-pushed
the
export-D93169309
branch
from
ae1f096 to
662a315
Compare
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
1 participant
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.
Summary:
This diff reworks the mslk nvfp4 stacked quantize kernel to hopefully be a bit simpler. As can be seen in gemm_ops.py, the new op minimizes extra artifacts needed for using the torch api for fp4fp4bf16_grouped_mm. This kernel is as performant as the mega kernel and hopefully robust, as shown in the added tests.
Reviewed By: jiawenliu64
Differential Revision: D93169309