Wentao YeandGitHub b5433b6f50 [Perf] Optimize dsv4 routing using specialized kernel, 2.94% E2E TPOT improvement (#48660) ...

Signed-off-by: yewentao256 <zhyanwentao@126.com>

2026-07-17 13:35:06 -07:00

..

marlin_moe_wna16

[11a/n] Migrate Marlin kernels to torch stable ABI (#45176)

2026-06-11 21:22:37 -07:00

permute_unpermute_kernels

[10c/n] Migrate MoE kernels to torch stable ABI (#44565)

2026-06-10 23:02:26 -07:00

dsv3_router_gemm_bf16_out.cu

[Misc] Use functions instead of PTX for the PDL instruction (#46984)

2026-07-01 19:38:35 -07:00

dsv3_router_gemm_entry.cu

New stable abi cleanup (#46656)

2026-07-03 14:02:26 +08:00

dsv3_router_gemm_float_out.cu

[Misc] Use functions instead of PTX for the PDL instruction (#46984)

2026-07-01 19:38:35 -07:00

grouped_topk_kernels.cu

New stable abi cleanup (#46656)

2026-07-03 14:02:26 +08:00

moe_align_sum_kernels.cu

add pad-aware reduce path (#48385)

2026-07-14 18:05:50 -07:00

moe_ops.h

add pad-aware reduce path (#48385)

2026-07-14 18:05:50 -07:00

moe_permute_unpermute_op.cu

New stable abi cleanup (#46656)

2026-07-03 14:02:26 +08:00

moe_wna16_utils.h

[10c/n] Migrate MoE kernels to torch stable ABI (#44565)

2026-06-10 23:02:26 -07:00

moe_wna16.cu

[10c/n] Migrate MoE kernels to torch stable ABI (#44565)

2026-06-10 23:02:26 -07:00

moeTopKFuncs.cuh

[10c/n] Migrate MoE kernels to torch stable ABI (#44565)

2026-06-10 23:02:26 -07:00

topk_softmax_kernels.cu

[Kernel] Applies routed_scaling_factor internally (#47408)

2026-07-07 02:00:54 -07:00

topk_softplus_sqrt_kernels.cu

[Perf] Optimize dsv4 routing using specialized kernel, 2.94% E2E TPOT improvement (#48660)

2026-07-17 13:35:06 -07:00

torch_bindings.cpp

add pad-aware reduce path (#48385)

2026-07-14 18:05:50 -07:00