[GH-ISSUE #335] Will DeepGEMM support GEMM + 1d1d quantization all in one fused kernel? #168

Open
opened 2026-10-26 21:14:04 +00:00 by bsakaguchi · 0 comments
Owner

Originally created by @TheTinyTeddy on GitHub (May 18, 2026).
Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/335

Many thanks for the great work!

I was wondering if DeepGEMM will support GEMM + optional epilogue (e.g. elementwise or norm) + 1d1d post quantization all in one fused kernel? This should greatly improve the performance by eliminating the input activation quantization kernel for the next GEMM OP.

Kind regards

Originally created by @TheTinyTeddy on GitHub (May 18, 2026). Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/335 Many thanks for the great work! I was wondering if DeepGEMM will support GEMM + optional epilogue (e.g. elementwise or norm) + 1d1d post quantization all in one fused kernel? This should greatly improve the performance by eliminating the input activation quantization kernel for the next GEMM OP. Kind regards
Sign in to join this conversation.
No labels
pull-request
No milestone
No project
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#168
No description provided.