Qi Yuhang d9def43dcd [Perf]Use Cooperative Schedule for H100 & H200 & H800 in fp8_blockwise_scaled_grouped_mm (#8722)

2025-08-02 21:13:47 -07:00

..

allreduce

[Feature] Integrate quick allreduce and select the best allreduce implementation (#6619)

2025-07-24 20:48:42 -07:00

attention

[bugifx] QWen-1M context support[2/3] using current cuda stream in the DCA's kernel for bugfix. (#8611)

2025-07-31 22:41:39 +08:00

cpu

Fix FP8 block quantization when N or K is not multiples of 128 (#8648)

2025-08-01 15:57:19 -07:00

cutlass_extensions

Update Cutlass in sgl-kernel to v4.1 (#8392)

2025-07-27 00:36:15 -07:00

elementwise

[AMD] Add silu_and_mul, gelu_and_mul, gelu_tanh_and_mul, and gelu_quick kernels for AMD GPUs (#7135)

2025-07-24 23:44:28 -07:00

gemm

fix per token cuda kernel hidden dim cannot divide by 16 (#8543)

2025-08-01 09:27:18 -07:00

grammar

…

kvcacheio

…

moe

[Perf]Use Cooperative Schedule for H100 & H200 & H800 in fp8_blockwise_scaled_grouped_mm (#8722)

2025-08-02 21:13:47 -07:00

spatial

[Bug] fix green context's incompatibility with cuda < 12.4 (#8701)

2025-08-02 15:23:11 -07:00

speculative

…

common_extension.cc

Revert "[1/2] sgl-kernel: Fuse routed scaling factor into select_experts" (#8706)

2025-08-02 20:14:30 +08:00

flash_extension.cc

…

torch_extension_rocm.cc

update sgl-kernel for EP: kernel part (#8514)

2025-07-30 22:19:55 -07:00