Michael Goin 34cd32fe30 [Perf][Kernel] Fused SiLU+Mul+Quant kernel for NVFP4 cutlass_moe (#31832) ...

Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>

2026-01-09 07:40:33 -07:00

..

awq

[Kernel] Fix awq error when n is not divisable by 128 (#13227)

2025-02-13 20:07:05 -08:00

cutlass_w4a8

[Kernel]Support W4A8 Grouped GEMM on Hopper (#29691)

2025-12-08 19:29:06 -08:00

fp4

[Perf][Kernel] Fused SiLU+Mul+Quant kernel for NVFP4 cutlass_moe (#31832)

2026-01-09 07:40:33 -07:00

fused_kernels

[Performance] Fused blockwise quant RMS norm (#27883)

2025-12-07 16:38:04 +00:00

gguf

[Bugfix][ROCm] Fix for warp_size uses on host (#21205)

2025-07-24 00:37:19 -07:00

gptq

[ROCm][GPTQ][Bugfix] Fix GPTQ GEMM kernel output zeroing race condition (#30719)

2025-12-29 01:13:14 -08:00

gptq_allspark

[Kernel][Quantization] add w4a8 support for marlin kernel (#24722)

2025-11-29 07:19:33 -08:00

gptq_marlin

[Kernel][Quantization][MoE] add marlin kernel support for turing (sm75) (#29901)

2025-12-16 14:35:28 -08:00

hadamard/hadacore

Use narrow over indexing in hadacore_transform to prep for ABI stable (#28756)

2025-11-15 01:10:15 -08:00

machete

Fix typos in comments across multiple files (#30345)

2025-12-09 20:05:28 -08:00

marlin/sparse

[Kernel/Quant] Remove the original marlin format and qqq (#23204)

2025-08-20 15:13:36 -04:00

w8a8

[MoE Refactor][3/N] Deprecate cutlass block quant fp8 (b200) (#30990)

2025-12-19 13:09:54 -08:00

activation_kernels.cu

[Performance][B200] silu_mul_quant: pack scales in int32 (#28358)

2025-11-13 10:16:55 -08:00

utils.cuh

[Feature][ROCm]Enable fusion pass for torch.compile on ROCm (#15050)

2025-03-31 04:42:18 -07:00

vectorization_utils.cuh

Make sure that vectorize_with_alignment produced vectorized global loads (#23182)

2025-08-21 20:06:54 +00:00

vectorization.cuh

[Perf] Tune scaled_fp8_quant by increasing vectorization (#18844)

2025-06-03 13:48:25 -07:00