pmanczakandPawel Manczak bdfb02674f [XPU][Kernel] Tune Triton W8A8 block-FP8 GEMM for Intel B70 (#56063) ...

Signed-off-by: Pawel Manczak <203671809+pmanczak@users.noreply.github.com>
Co-authored-by: Pawel Manczak <203671809+pmanczak@users.noreply.github.com>

2026-10-02 12:56:11 +01:00

..

cpu

[Model] Add DeepSeek-V4 CPU backend (#55355)

2026-09-10 14:14:48 +08:00

deepgemm

…

ir

[Benchmark] Retire stale benchmarks and consolidate RMSNorm (#57083)

2026-09-16 13:21:43 -04:00

__init__.py

…

bench_concat_mla_q.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

bench_cp_gather_fp8.py

…

bench_cpu_sampling.py

[CPU][Perf] Add vectorized Sampler Kernel (#53913)

2026-09-30 15:16:28 +08:00

benchmark_2d_silu_mul_fp8_quant.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_activation.py

…

benchmark_block_fp8_gemm.py

…

benchmark_cp_gather.py

…

benchmark_cutlass_moe_fp8.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_cutlass_moe_nvfp4.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_device_communicators.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_dsv41_mega_attn.py

[Model][DSv4.1] FlashMLA mega attention and the NVFP4 compressed KV cache (#56935)

2026-09-16 03:07:50 -07:00

benchmark_flydsl_moe_w4a16.py

…

benchmark_fp8_gemm.py

…

benchmark_fused_collective.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_fused_moe_lora_one_shot.py

…

benchmark_fused_q_cutedsl.py

…

benchmark_fused_topk.py

…

benchmark_gelu_and_mul_sparse.py

[Performance] Add Triton kernel for Gemma3n sparse GELU (#48498)

2026-09-14 07:42:15 +00:00

benchmark_gemm_rs_ar.py

[Kernel][DSV4.1] Fuse MXFP8 wo_b GEMM with sequence-parallel reduce-scatter (#57428)

2026-09-22 01:07:17 -07:00

benchmark_grouped_gemm_cutlass.py

…

benchmark_host_buffer_kv_copy.py

[KV Connector][NIXL] Coalesce host-buffer KV copies across cache groups (#54483)

2026-10-01 13:10:54 -05:00

benchmark_hy_v4_ihc.py

…

benchmark_inkling_qkvr_prep.py

…

benchmark_int8_gemm.py

…

benchmark_k3_cutedsl_residual.py

…

benchmark_kimi_k3_attn_res.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_kimi_k3_kda_decode.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_kimi_k3_kda_projection.py

…

benchmark_kimi_k3_latent_moe_tail.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_kimi_k3_sp_collectives.py

…

benchmark_layernorm.py

…

benchmark_lora.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_machete.py

…

benchmark_marlin.py

[Kernel] Remove AllSpark INT8 W8A16 GEMM backend (#58001)

2026-09-22 14:07:34 -07:00

benchmark_mla_k_concat.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_moe_align_block_size.py

[Kernel][MoE] Optimize batched_moe_align_block_size with cooperative writes (#53280)

2026-09-11 12:05:22 -07:00

benchmark_moe_defaults.py

[CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)

2026-09-16 10:57:24 -07:00

benchmark_moe_permute_unpermute.py

…

benchmark_moe.py

…

benchmark_mrope.py

…

benchmark_multimem_reduce_scatter.py

[Perf][Distributed] Add low-SM multimem reduce-scatter for SM100/SM103 (#55072)

2026-09-25 03:27:21 +00:00

benchmark_mxfp4_qutlass.py

…

benchmark_nvfp4_gemm.py

…

benchmark_nvfp4_quant.py

…

benchmark_nvfp4_qutlass.py

…

benchmark_paged_attention.py

…

benchmark_per_token_group_quant.py

[Kernel][Perf] Register-resident path for per-token-group 8-bit quant (#55330)

2026-09-25 08:11:42 -07:00

benchmark_per_token_quant_fp8.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_persistent_topk.py

[Perf][Kernel] Add sampled filtering for persistent top-k (#56346)

2026-09-15 14:15:59 -07:00

benchmark_quant.py

…

benchmark_rdna_hybrid_w4a16_gemm.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_relu_squared.py

…

benchmark_reshape_and_cache_flash.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_reshape_and_cache.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_rope.py

…

benchmark_router_gemm.py

…

benchmark_selective_state_update.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_shapes.py

…

benchmark_silu_mul_fp8_quant.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_thinking_budget.py

…

benchmark_top_k_per_row_decode.py

[ROCm][Perf][GLM-5.3-Flash] Add AITER topk backend for decodes (#58167)

2026-10-02 11:47:41 +01:00

benchmark_trtllm_decode_attention.py

…

benchmark_trtllm_prefill_attention.py

…

benchmark_vit_aiter_fp8_attn.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

benchmark_vit_bilinear_pos_embed.py

…

benchmark_vit_fp8_attn.py

…

benchmark_vocab_parallel_embedding.py

[CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)

2026-09-16 10:57:24 -07:00

benchmark_w8a8_block_fp8.py

[XPU][Kernel] Tune Triton W8A8 block-FP8 GEMM for Intel B70 (#56063)

2026-10-02 12:56:11 +01:00

graph_machete_bench.py

…

requirements.txt

…

utils.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

weight_shapes.py

…