deepseek-ai/DeepGEMM
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepGEMM.git synced 2026-11-05 00:47:26 +00:00
Code Issues 162 Projects Packages Wiki Activity

[GH-ISSUE #50] "it employs CUDA-core two-level accumulation (promotion)" #32

New issue
Closed
opened 2026-10-26 21:10:55 +00:00 by bsakaguchi · 1 comment
bsakaguchi commented 2026-10-26 21:10:55 +00:00
Owner
Copy link

Originally created by @cxw-droid on GitHub (Mar 7, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/50

May I know what is the meaning of the two-level here? Does it mean FP8->BF16->FP32? Thanks.

Originally created by @cxw-droid on GitHub (Mar 7, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/50 May I know what is the meaning of the `two-level` here? Does it mean FP8->BF16->FP32? Thanks.
bsakaguchi commented 2026-10-26 21:10:57 +00:00
Author
Owner
Copy link

@soundOfDestiny commented on GitHub (Mar 8, 2025):

In the current Tensor Core
implementation of the NVIDIA Hopper architecture, FP8 GEMM (General Matrix Multiply)
employs fixed-point accumulation, aligning the mantissa products by right-shifting based on
the maximum exponent before addition. Our experiments reveal that it only uses the highest 14 bits of each mantissa product after sign-fill right shifting, and truncates bits exceeding this range.

FP8 GEMM -> FP22 result per 128 along K dimension -> FP32 accumulation

<!-- gh-comment-id:2707877416 --> @soundOfDestiny commented on GitHub (Mar 8, 2025): In the current Tensor Core implementation of the NVIDIA Hopper architecture, FP8 GEMM (General Matrix Multiply) employs fixed-point accumulation, aligning the mantissa products by right-shifting based on the maximum exponent before addition. Our experiments reveal that it only uses the highest 14 bits of each mantissa product after sign-fill right shifting, and truncates bits exceeding this range. FP8 GEMM -> FP22 result per 128 along K dimension -> FP32 accumulation
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
public-sync-20261014
0930-release
nv_dev
fix-mega-moe-fence
public-release-2609
public-release-2607
public-release-2606
mega-update
public-release-260416
ko3n1g/ci/build-wheels
nv_dev_f8e8fb5
nv_dev_950d31a
nv_dev_67fc648
nv_dev_c491439
nv_dev_4ff3f54
v2.1.1.post3
v2.1.1.post2
v2.1.1.post1
v2.1.1
v2.1.0
Labels
Clear labels   
pull-request
Mirrored from GitHub Pull Request

No labels
pull-request
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/DeepGEMM#32
No description provided.