Skip to content

[None][perf] GLM-5.3-Flash C++ kernel speedups: vectorized FP8 1x128 quant and 288-expert router GEMM - #20390

Merged
ruocheng-nv merged 3 commits into
NVIDIA:mainfrom
ruocheng-nv:perf/glm5.3-flash-cpp-kernels
Oct 19, 2026
Merged

[None][perf] GLM-5.3-Flash C++ kernel speedups: vectorized FP8 1x128 quant and 288-expert router GEMM#20390
ruocheng-nv merged 3 commits into
NVIDIA:mainfrom
ruocheng-nv:perf/glm5.3-flash-cpp-kernels

[None][fix] Pack router GEMM operands for both dispatch paths

8d6abe2
Select commit
Loading
Failed to load commit list.

Workflow runs completed with no jobs