[1/n]: add cutlass W4A8 moe kernel for hopper architecture #7772

yangsijia-serena · 2025-07-04T08:06:43Z

Motivation

This is the sgl-kernel part of #7762. A series of PRs will enable running DeepSeek-R1-W4AFP8 using sglang.

Additionally, this kernel supports other MoE models with INT4 MoE weight and FP8 activation quantization.

Co-author: yicwang yichen.wang@bytedance.com

Modifications

In this kernel implementation, we focus on two main areas:
1. Function Implementation:

scaled_mm_entry.cu: This serves as the entry point for the CUTLASS W4A8 MoE.
w4a8_moe_data.cu: Contains the logic for computing expert_offsets and problem_size
w4a8_grouped_mm_c3x.cu:Responsible for dispatching grouped GEMM operations.
w4a8_grouped_mm_c3x.cuh: Contains the key computation logic.
cutlass_extensions/: This section is adapted from trtllm and contains the separate CUTLASS components needed for the W4A8 grouped GEMM.

2. Tailored Performance Tuning for the DeepSeek R1 Model:

We explore the optimal CUTLASS template configuration for the MoE layer of DeepSeek R1 to maximize performance.
The tuning results are integrated into the dispatch logic in w4a8_grouped_mm_c3x.cu

Checklist

Format your code according to the Code Formatting with Pre-Commit.
Add unit tests as outlined in the Running Unit Tests.
Update documentation / docstrings / example tutorials as needed, according to Writing Documentation.
Provide throughput / latency benchmark results and accuracy evaluation results as needed, according to Benchmark and Profiling and Accuracy Results.
For reviewers: If you haven't made any contributions to this PR and are only assisting with merging the main branch, please remove yourself as a co-author when merging the PR.
Please feel free to join our Slack channel at https://slack.sglang.ai to discuss your PR.

Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com>

gemini-code-assist

Summary of Changes

Hello @yangsijia-serena, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request introduces a highly optimized CUTLASS-based kernel for Mixture of Experts (MoE) operations on NVIDIA Hopper architecture (SM90). It specifically targets efficient execution of models using INT4 weights and FP8 activations, such as DeepSeek-R1-W4AFP8, by leveraging advanced grouped GEMM capabilities and tailored performance tuning. This is the first in a series of PRs to enable quantized MoE models in sglang.

Highlights

New W4A8 MoE Kernel: Implemented a new CUDA kernel for W4A8 (Weight 4-bit, Activation 8-bit) Mixture of Experts (MoE) operations, specifically designed for NVIDIA Hopper GPUs (SM90).
CUTLASS 3.x Integration: Leverages CUTLASS 3.x for efficient grouped GEMM, including specialized components for handling mixed-precision inputs (INT4 weights and FP8 activations). This involves new files like mixed_input_utils.hpp and sm90_mma_array_tma_gmma_rs_warpspecialized_mixed_input_.hpp.
Core Kernel Components: Added scaled_mm_entry.cu as the entry point, w4a8_moe_data.cu for expert data preparation, and w4a8_grouped_mm_c3x.cu/.cuh for the core grouped GEMM dispatch and computation.
Python Bindings: Exposed the new kernel functionalities (get_cutlass_w4a8_moe_mm_data and cutlass_w4a8_moe_mm) to the Python API for easy integration, defined in the new sgl_kernel/python/sgl_kernel/cutlass_moe.py module.
Performance Tuning: Includes specific CUTLASS template configurations optimized for the DeepSeek R1 MoE layer to maximize performance.
Build System Updates: Integrated new CUDA source files into the CMakeLists.txt and updated common_extension.cc and sgl_kernel_ops.h for proper compilation and function registration.
Unit Tests: Added a new test file (test_cutlass_w4a8_moe_mm.py) to validate the correctness and functionality of the new INT4/FP8 grouped GEMM kernel, including int4 packing and scale interleaving.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces a new CUTLASS W4A8 MoE kernel tailored for the Hopper architecture, which is a significant addition for enabling mixed-precision MoE models. The changes involve adding new CUDA/C++ source files, integrating them into the CMake build system, and exposing them via PyTorch bindings. The implementation leverages advanced CUTLASS features and CUDA programming patterns for performance. The review identified several areas for improvement, including addressing potential division-by-zero issues with input parameters, clarifying complex logic, and enhancing code readability by replacing magic numbers with named constants.

gemini-code-assist · 2025-07-04T08:09:03Z

...tlass_extensions/gemm/collective/sm90_mma_array_tma_gmma_rs_warpspecialized_mixed_input_.hpp

+    int chunk_size = 0;
+    ElementZero const** ptr_Z = nullptr;


The chunk_size member of Params is initialized to 0, and reload_factor is calculated using chunk_size. If chunk_size remains 0, this will lead to a division-by-zero error. Ensure chunk_size is always set to a valid non-zero value before reload_factor is computed or used.

gemini-code-assist · 2025-07-04T08:09:03Z

...tlass_extensions/gemm/collective/sm90_mma_array_tma_gmma_rs_warpspecialized_mixed_input_.hpp

+        // Temporary factor which will determine which k tile to reload from gmem. Needed so we don't modify tma
+        // transaction bytes on the fly. We must do a ceiling divide here to correctly handle with chunk_size == K. In
+        // that case, we don't require that K is a multiple of the threadblock tile K
+        const int scale_load_k = *k_tile_iter / 1;


The calculation const int scale_load_k = *k_tile_iter / 1; is redundant. If reload_factor (commented out on line 850) was intended to be used here, the current line is incorrect. Please clarify the intended logic for scale_load_k and correct the calculation if necessary.

gemini-code-assist · 2025-07-04T08:09:03Z

sgl-kernel/csrc/moe/cutlass_moe/w4a8/w4a8_grouped_mm_c3x.cu

+  using KernelSchedule = cutlass::gemm::KernelPtrArrayTmaWarpSpecializedCooperative;
+  using EpilogueSchedule = cutlass::epilogue::PtrArrayTmaWarpSpecializedCooperative;
+
+  uint32_t const m = a_tensors.size(0) / topk;


The calculation uint32_t const m = a_tensors.size(0) / topk; involves division by topk. If topk can be zero, this will result in a division-by-zero runtime error. Given that the Python wrapper cutlass_w4a8_moe_mm has a default topk=0, this is a potential issue. Add a check to ensure topk is greater than zero before performing this division.

gemini-code-assist · 2025-07-04T08:09:03Z

sgl-kernel/python/sgl_kernel/cutlass_moe.py

+    chunk_size: int = 0,
+    topk: int = 0,


The chunk_size and topk parameters have default values of 0. As noted in w4a8_grouped_mm_c3x.cu, topk is used in a division operation, and chunk_size is used in calculations that could lead to division by zero if it remains 0. It is recommended to either provide non-zero default values or add runtime checks to prevent division by zero.

consider non-zero default value or add checks here

gemini-code-assist · 2025-07-04T08:09:03Z

...tlass_extensions/gemm/collective/sm90_mma_array_tma_gmma_rs_warpspecialized_mixed_input_.hpp

+#include "cutlass/trace.h"
+#include "cutlass_extensions/detail/collective/mixed_input_utils.hpp"
+
+#define GROUP_SIZE 128


The GROUP_SIZE is defined as a macro. For better type safety, scope, and debugging, consider defining it as a static constexpr member of the relevant class or a global static constexpr variable in an anonymous namespace.

Suggested change

#define GROUP_SIZE 128

static constexpr int GROUP_SIZE = 128;

gemini-code-assist · 2025-07-04T08:09:04Z

sgl-kernel/csrc/cutlass_extensions/detail/collective/mixed_input_utils.hpp

+  template <class EngineIn, class EngineOut, class LayoutIn, class LayoutOut, class... Ts>
+  CUTLASS_DEVICE static void convert_A_kblock(
+      Tensor<EngineIn, LayoutIn> const& tCrA_load, Tensor<EngineOut, LayoutOut>& tCrA_mma, int const k_block) {
+    static_assert(is_rmem<EngineIn>::value, "Input tensor for A conversion must come from registers");
+    static_assert(is_rmem<EngineOut>::value, "Output tensor for A conversion must come from registers");
+    static_assert(cosize_v<LayoutIn> == cosize_v<LayoutOut>);
+    static_assert(size_v<LayoutIn> == cosize_v<LayoutIn>);
+    static_assert(size_v<LayoutOut> == cosize_v<LayoutOut>);
+    using SrcType = typename EngineIn::value_type;
+
+    Tensor src = tCrA_load(_, _, k_block);
+    Tensor dst = tCrA_mma(_, _, k_block);
+
+    CUTE_STATIC_ASSERT_V(
+        size(src(_, 0)) == cosize(src(_, 0).layout()), "The first mode of tensor src must be contiguous in memory");
+    // try to make the size of the first mode equal to 32bit
+    int constexpr NumValPerSrcReg = cute::min(decltype(size(src(_, 0)))::value, ceil_div(32, sizeof_bits_v<SrcType>));
+    Tensor src_vm = cute::group_modes<1, -1>(cute::zipped_divide(src, Int<NumValPerSrcReg>{}));
+    Tensor dst_vm = cute::group_modes<1, -1>(cute::zipped_divide(dst, Int<NumValPerSrcReg>{}));
+
+    // KernelConversionMode == ConversionMode::DirectConvert
+    CUTLASS_PRAGMA_UNROLL
+    for (int i = 0; i < size<1>(dst_vm); ++i) {
+      LayoutAwareConvert(src_vm(_, i), dst_vm(_, i));
+    }
+  }


The convert_A_kblock function appears to be a specialized version of dequantize_A_kblock for ConversionMode::DirectConvert. Consider unifying these two functions to reduce code duplication and improve maintainability, as dequantize_A_kblock already handles the DirectConvert case.

gemini-code-assist · 2025-07-04T08:09:04Z

sgl-kernel/python/sgl_kernel/cutlass_moe.py

+    d_strides: torch.tensor,
+    s_strides: torch.tensor,
+    chunk_size: int = 0,
+    topk: int = 0,


The type hints for experts_offsets and problem_sizes are torch.tensor. For clarity and consistency with PyTorch's API, consider using torch.Tensor instead.

Suggested change

d_strides: torch.tensor,

s_strides: torch.tensor,

chunk_size: int = 0,

topk: int = 0,

experts_offsets: torch.Tensor,

problem_sizes: torch.Tensor,

a_strides: torch.Tensor,

b_strides: torch.Tensor,

gemini-code-assist · 2025-07-04T08:09:04Z

sgl-kernel/csrc/moe/cutlass_moe/w4a8/w4a8_moe_data.cu

+
+#include <iostream>
+
+constexpr uint64_t THREADS_PER_EXPERT = 512;


The THREADS_PER_EXPERT is defined as a constexpr macro. For better maintainability and to avoid potential issues with macro expansion, consider defining it as a static constexpr variable.

static constexpr uint64_t THREADS_PER_EXPERT = 512;

AniZpZ · 2025-07-04T09:35:51Z

sgl-kernel/python/sgl_kernel/cutlass_moe.py

+    chunk_size: int = 0,
+    topk: int = 0,


consider non-zero default value or add checks here

AniZpZ · 2025-07-04T09:53:21Z

sgl-kernel/tests/test_cutlass_w4a8_moe_mm.py

+        )
+        print(f"    AssertionError: {e}")
+        raise
+


consider add multi expert unittest
@pytest.mark.parametrize("batch_size,k,n,num_experts", [
(4, 384, 768, 4),
(8, 512, 1024, 8),
])
def test_int4_fp8_grouped_gemm_multi_experts(batch_size, k, n, num_experts):

fixed and added~

…e_mm. Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com>

Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com>

…ct#7772) Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com>

@mickqian

* Use seq_len_fill_value in the cuda graph runners (sgl-project#7233) * support custom weight loader for model runner (sgl-project#7122) Co-authored-by: kavioyu <kavioyu@tencent.com> * Fix AMD speculative decoding (sgl-project#7252) * [Refactor] OAI Server components (sgl-project#7167) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * OAI Server Skeleton & Core Utility Endpoints (sgl-project#7179) * [amd] Opt dsv3 moe (sgl-project#7160) Co-authored-by: wunhuang <wunhuang@amd.com> * update ci node for xeon (sgl-project#7265) * feat: mtp support dp-attention (sgl-project#6081) Co-authored-by: austindeng <austindeng@tencent.com> Co-authored-by: tianqilin.99 <tianqilin.99@bytedance.com> Co-authored-by: Qiaolin Yu <liin1211@outlook.com> Co-authored-by: ch-wan <cwan39@gatech.edu> * support qwen2 running on ascend npu device (sgl-project#7022) Co-authored-by: 刁莹煜 <diaoyingyu1@hisilicon.com> * Fix Deepseek R1 0528 FP4 tensor name mismatch issue during weights loading. (sgl-project#7164) * bugfix(tool call ebnf): Fix EBNF generation for optional function parameters (sgl-project#7283) * Fix AWQ Dequant and Weight Loading of deepseek v2 (sgl-project#6842) * fix: resolve b200 dsv3 mtp issue (sgl-project#7286) * ci: Fix test_ebnf_generate_all_optional_function_params (sgl-project#7288) * fix: only enable flash_attn test on sm80 sm90 (sgl-project#7289) * [PD] Support get local ip from NIC for PD disaggregation (sgl-project#7237) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * [PD] Add custom memory pool option to support Mooncake PD with NVLink (sgl-project#7264) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * Upstreaming hicache bug fixes (sgl-project#7267) * Update python API of activation, topk, norm and rope and remove vllm dependency (sgl-project#6614) Co-authored-by: Wu, Chunyuan <chunyuan.wu@intel.com> Co-authored-by: jianan-gu <jianan.gu@intel.com> Co-authored-by: sdp <sdp@gnr799219.jf.intel.com> * Fix hicache benchmark script bug - some sampled input_request is [] (sgl-project#7300) * chore: change logs from`INFO` to `DEBUG` for dp and add force quit for tokenizer manager (sgl-project#7251) * update invalid link in doc (sgl-project#7297) * Fix mini_lb for PD with long output: limit chunk size of decode response (sgl-project#7301) Signed-off-by: ch-tiger1 <xyz@ch-tech.ip-ddns.com> Co-authored-by: ch-tiger1 <xyz@ch-tech.ip-ddns.com> * Fix profiler error when there are idle passes (sgl-project#7003) * [pd] optimize dockerfile for pd disaggregation (sgl-project#7319) Co-authored-by: zhyncs <me@zhyncs.com> * Merge PDLB (Prefill-Decode Load Balancer) into SGLang Router (sgl-project#7096) * Add more refactored openai test & in CI (sgl-project#7284) * fix: resolve blackwell deepep image issue (sgl-project#7331) * add seed in CPU UTs to avoid flaky failure (sgl-project#7333) * Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (sgl-project#7099) * Reintroduce tiny fix sampler error when prob is not contiguous (sgl-project#7354) * [Refactor] Clean up radix cache related API (sgl-project#7303) Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu> * Put `_normalize_rid` before other normalization in `io_struct` (sgl-project#7363) * [PD] Transfer hidden states for mtp when disaggregation (sgl-project#7242) * [Bugfix][PD] Set conclude state before clear when failure happens (sgl-project#7362) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * docs: update installation (sgl-project#7366) * [Docker] optimize dockerfile remove deepep and blackwell merge it to… (sgl-project#7343) Co-authored-by: Yineng Zhang <me@zhyncs.com> * Clean unused import for mimo mtp model (sgl-project#7370) * [Bugfix]Fix hang bug using dp attention with HiRadixCache (sgl-project#7159) Signed-off-by: huanglong <huanglong@linux.alibaba.com> * [Doc] add embedding rerank doc (sgl-project#7364) * Fix judgment condition for enabling Deepseek V3/R1 shared expert fusion optimization (sgl-project#7371) * Feat/refactor embedding server (sgl-project#7322) * Purge VerlEngine (sgl-project#7326) Signed-off-by: Ata Fatahi <immrata@gmail.com> * support return logprobs for pipeline (sgl-project#7356) Co-authored-by: Zhang Kaihong <zhangkaihong.zkh@alibaba-inc.com> * [PD] Optimize custom mem pool usage and bump mooncake version (sgl-project#7393) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * Support THUDM/GLM-4-0414 (GLM-Z1) Glm4ForCausalLM architecture. (sgl-project#5485) * Refine OpenAI serving entrypoint to remove batch requests (sgl-project#7372) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> Co-authored-by: Chang Su <csu272@usc.edu> * [Feature] Comprehensive Hybrid Parallelism Support (sgl-project#6389) * [DeepSeekNextN] fix: residual of head norm can be None (sgl-project#7398) * [OAI refactor] Add rerank and score serving (sgl-project#7399) Co-authored-by: Chang Su <chang.s.su@oracle.com> * [OAI Server Refactor] [ChatCompletions & Completions] Implement UsageInfo Processor (sgl-project#7360) Co-authored-by: Chang Su <chang.s.su@oracle.com> * Fix All-Gather under world size one (sgl-project#7219) * Optimize DP attn scheduling for speculative decoding (sgl-project#7285) * Update usage_processor.py (sgl-project#7402) * Fix 7285 Merge Conflicts (sgl-project#7403) * chore: upgrade mooncake-transfer-engine 0.3.4 (sgl-project#7401) * [OAI Server Refactor] [ChatCompletions & Completions] Support Return Hidden State (sgl-project#7329) Signed-off-by: keru <rukeyang@gmail.com> * Remove batches api in docs & example (sgl-project#7400) * [BugFix]: fix EmbeddingReqInput single input error (sgl-project#7396) * [BugFix]fix qwen25 invoke function call streaming responses with curly braces as the starting indicator (sgl-project#7394) * fix overlap pagecount (sgl-project#6984) Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu> * fix: Fix CI test_function_call_parser.py (sgl-project#7425) * Fix CPU offloading for MLA memory pool (sgl-project#7409) * [fix] PD disaggregation when enable mtp and tp!=dp (sgl-project#7420) * feat(oai refactor): Replace `openai_api` with `entrypoints/openai` (sgl-project#7351) Co-authored-by: Jin Pan <jpan236@wisc.edu> * Refactor LoRAManager and LoRAMemoryPool state management logic for dynamic LoRA loading support (sgl-project#7412) * refactor(test): reorganize OpenAI test file structure (sgl-project#7408) * [minor] simplify the `TokenToKVPoolAllocator` (sgl-project#7414) * Tiny add logging for GC (sgl-project#7406) * FlashInfer NVFP4 MoE with EP & 2-stream shared expert (sgl-project#7327) Co-authored-by: JieXin Liang <Alcanderian@users.noreply.github.com> Co-authored-by: alcanderian <alcanderian@gmail.com> * Remove copy after bmm (sgl-project#7441) * Fix torch compile run (sgl-project#7391) Co-authored-by: wunhuang <wunhuang@amd.com> Co-authored-by: Sai Enduri <saimanas.enduri@amd.com> * [misc] Add PD service discovery support in router (sgl-project#7361) * add fused moe config for qwen3 in triton3.3.1 (sgl-project#7445) * Fix CUDA Graph Check under Deepep with DP FFN (sgl-project#7451) * Update hyperparameter_tuning.md (sgl-project#7454) * feat: integrate deepgemm into EPMoE (sgl-project#6821) Co-authored-by: tianqilin.99 <tianqilin.99@bytedance.com> Co-authored-by: TianQiLin666666 <1834987979@qq.com> Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com> * Solve docker build failed in the virtual machine (sgl-project#7290) Co-authored-by: wunhuang <wunhuang@amd.com> Co-authored-by: Sai Enduri <saimanas.enduri@amd.com> Co-authored-by: HAI <hixiao@gmail.com> * Fix a bug in BatchTokenIDOut & Misc style and dependency updates (sgl-project#7457) * [CI] Upgrade mooncake to 0.3.4.post1 to fix 8 gpu tests (sgl-project#7472) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * Fix prefill OOM due to wrong token calculation when page > 1 (sgl-project#7397) * feat(func_call): Add more check in `BaseFormatDetector.parse_streaming_increment` (sgl-project#7479) * Fix dtype for idle input in spec decoding (sgl-project#7456) * update mooncake in dockerfile (sgl-project#7480) * kvcache io kernels and test case (sgl-project#7382) * [perf] slightly imporve DeepSeek-R1-FP4 TP8 (sgl-project#7481) * Quick fix for DeepGemm requant to also cover MTP. (sgl-project#7378) * Support weight loading without mmap (sgl-project#7469) * ci: Revert openai_server related tests in AMD suites (sgl-project#7449) * Perormance: Enable cuda graph for dp idle batch (sgl-project#7269) Co-authored-by: austindeng <austindeng@tencent.com> Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com> Co-authored-by: ch-wan <cwan39@gatech.edu> * bugfix: Prevent global mutation of conv.stop_str across requests (sgl-project#7347) Co-authored-by: Chang Su <chang.s.su@oracle.com> * Fix RequestValidationError response format (sgl-project#7487) * Fix MTP with Deepseek R1 Fp4 (sgl-project#7376) * chore: bump sgl-kernel v0.2.0 (sgl-project#7490) * chore: bump v0.4.8 (sgl-project#7493) * [AMD] add aiter fused moe in DeepEP path (sgl-project#7268) * enable aiter_biased_grouped_topk kernel (sgl-project#7423) * [PD Disaggregation] replace transfer with batch transfer for better performance (sgl-project#7236) * Remove cumsum_buffer initilization (sgl-project#7439) * [benchmark] fbgemm benchmark support bandwidth report and support fbgemm_cutlass_gmm (sgl-project#7422) * Support multi-thread model weight loading (sgl-project#7277) * [PD] NIXL: Register kv args in advance and cleanup finished requests (sgl-project#6717) * fix: Add `--model` as an alias for `--model-path` in server_args (sgl-project#7505) * misc: Improvement to serving_chat.py and add more ut (sgl-project#7489) * Fuse sorted_token_ids padding to moe_align_block_size kernel (sgl-project#7437) * [OAI] patch origin request_id logic (sgl-project#7508) * [PD][Spec] Fix hidden state transfer for spec decode (sgl-project#7516) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * EPLB support for MTP (sgl-project#7510) * clean duplicate code (sgl-project#7512) * [ci] add router benchmark script and CI (sgl-project#7498) * fix: force synchronization between TP workers when update_weights (sgl-project#6626) Co-authored-by: dangkai.dk <dangkai.dk@alibaba-inc.com> * [CPU] [BF16] Call fused_experts_cpu, weight_packed_linear and bmm_cpu kernel in DeepSeek model (sgl-project#6641) Co-authored-by: Thien Tran <gau.nernst@yahoo.com.sg> * [CI] Upgrade mooncake to v0.3.4.post2 to fix potential slice failed bug (sgl-project#7522) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * npu fused op (sgl-project#7386) Co-authored-by: Li Junwen <lijunwen13@hisilicon.com> * feat: send kvmetrics from sglang scheduler (sgl-project#6721) * [PD] Add different TP sizes support for no-MLA models (sgl-project#6793) Co-authored-by: shangmingc <csmthu@gmail.com> Co-authored-by: Shangming Cai <caishangming@linux.alibaba.com> * enable aiter fp8 blockscale quant (sgl-project#7520) * take aiter get_rope back (sgl-project#7521) * Fix typo of flash_cache (sgl-project#7513) * feat: add return hidden_states at async generation (sgl-project#7507) * minor: 'role' must be system/assistant/tool, but case insensitive for now (sgl-project#7499) * Fix FP8 KV Cache Support in FA3 Backend (sgl-project#7148) * Fix gathered_buffer issues in tbo (sgl-project#7531) * [PD] Raise error for incompatible mooncake version and some minor fixes (sgl-project#7527) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * [CMake] Fix sgl-kernel CMakeLists for Blackwell (sgl-project#7543) * Add Tencent HunYuanMoEV1 model support (sgl-project#7549) * Update seed in CPU UTs to avoid flaky failure with single test (sgl-project#7544) * chore: improve ci bug reporting (sgl-project#7542) * chore: remove vlm unnecessary import (sgl-project#7541) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> Co-authored-by: yhyang201 <yhyang201@gmail.com> Co-authored-by: Mick <mickjagger19@icloud.com> * chore: bump v0.4.8.post1 (sgl-project#7559) * [PD][NIXL] Set is_sorted=False to fix NIXL_ERR_NOT_FOUND (sgl-project#7330) * [Fix] incorrect assert in EPLB (sgl-project#7575) * Updates Gemma3n MLP layer to adapt latest transformers version (sgl-project#7573) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * Fix MTP error when enabling two-batch overlap (sgl-project#7569) * Add e2e test for multi instance multi stage memory release/resume occupuation (sgl-project#7208) Signed-off-by: Ata Fatahi <immrata@gmail.com> * [CI] Add CI Testing for Prefill-Decode Disaggregation with Router (sgl-project#7540) * Updates transformers and timm dependencies (sgl-project#7577) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * feat: support compatibility between MTP and two-batch-overlap (sgl-project#7225) Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com> * Move multimodal processors into a separate folder (sgl-project#7581) * Fix broken CI TestVILAServer (sgl-project#7610) * [router] add centralized configuration module for sgl-router (sgl-project#7588) * Fix: Minicpm (sgl-project#7612) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * Hybrid kv cache for LLaMA4 (sgl-project#6563) Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com> Co-authored-by: tarinkk <rt572@physics.rutger.edu> Co-authored-by: tarinkk <rt572@rutgers.physics.edu> Co-authored-by: Hanming Lu <69857889+hanming-lu@users.noreply.github.com> * [CPU] add optimizations for INT8 and FP8 DeepSeek (sgl-project#6769) Co-authored-by: Zheng, Beilei <beilei.zheng@intel.com> * Tiny add logs for expert location updater (sgl-project#7308) * Fix flakiness in LoRA batch test. (sgl-project#7552) * [BUG] fix local_rank in initialize_dp_attention (sgl-project#7584) * Support dynamic LoRA loading / unloading in engine/server API (sgl-project#7446) * [PD] Respect sampling_params.max_new_tokens when PD disaggregation is activated (sgl-project#7598) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * fix unit tests (sgl-project#7618) * Let ep_scatter support arbitrary strides / ue8m0 format (sgl-project#7309) * Let EP prefill support new DeepGEMM (sgl-project#7310) * docs: add gb200 nvl72 and a16z grant (sgl-project#7620) * oai: Adds support for OpenAI chat completions API in bench_serving (sgl-project#7036) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> Co-authored-by: yhyang201 <47235274+yhyang201@users.noreply.github.com> Co-authored-by: Mick <mickjagger19@icloud.com> * [bugfix] Remove PR comment posting from Rust benchmark workflow (sgl-project#7625) * [Minor] clean up multimodal processor and tokenizer manager (sgl-project#7624) * Add dsv3 fused a gemm to sgl-kernel (sgl-project#7630) * Add @mickqian as the CODEOWNERS of multimodal (sgl-project#7636) * Fix stream reasoning parser and Adds Kimi reasoning parser (sgl-project#7432) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * Fix sgl-router startup crash (sgl-project#7619) * [bugfix] fix runtime dropping panic in editable (sgl-project#7628) * Move files related to EPLB (sgl-project#7580) * [misc] reduce weird rope_scaling_factor warning (sgl-project#7176) * [AMD] Add unit-test-sgl-kernel-amd to AMD CI (sgl-project#7539) * Update CODEOWNERS (sgl-project#7640) * [EAGLE] remove a wrong adjustment for page_size > 1 & topk > 1 in server_args.py (sgl-project#7643) * [CPU] add c++ kernel to bind CPU cores and memory node (sgl-project#7524) * Improve streaming, log_level, memory report, weight loading, and benchmark script (sgl-project#7632) Co-authored-by: Kan Wu <wukanustc@gmail.com> * Add dsv3 router gemm kernel (sgl-project#7627) * chore: upgrade flashinfer v0.2.7 jit (sgl-project#7663) * [doc] update lws doc for pd (sgl-project#7318) * Fix: sync prepare_fp8_layer_for_marlin with latest vllm changes (sgl-project#7648) * Add small requirements for benchmark/parse_result tools (sgl-project#7671) * [CPU] remove process_group from inputs of shm_allreduce and shm_allgather (sgl-project#7486) * chore: bump sgl-kernel v0.2.1 (sgl-project#7675) * support llama4 eagle3 (sgl-project#6985) Co-authored-by: shuaills <shishuaiuoe@gmail.com> Co-authored-by: Shenggui Li <somerlee.9@gmail.com> Co-authored-by: Yingyi Huang <yingyihuang2000@outlook.com> Co-authored-by: yizhang2077 <1109276519@qq.com> * Refactor mm processors and Enable mixed modality processing (sgl-project#7629) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * upgrade sgl kernel to 0.2.1 for main (sgl-project#7676) * add description for llama4 eagle3 (sgl-project#7688) * fix(model loader): use safe_open to prevent file handle leaks. (sgl-project#7684) * chore: upgrade flashinfer v0.2.7.post1 (sgl-project#7698) * Improve error handling for requests with unloaded LoRA path(s) (sgl-project#7642) * Apply dsv3_fused_a_gemm kernel (sgl-project#7635) * Fix GPTQMarlinMoE (sgl-project#7697) * [1/n] apply wna16marlin kernel in moe weight only quantization (sgl-project#7683) Co-authored-by: 晟海 <huangtingwei.htw@antgroup.com> Co-authored-by: yych0745 <1398089567@qq.com> Co-authored-by: HandH1998 <1335248067@qq.com> Co-authored-by: 弋云 <yiyun.wyt@antgroup.com> Co-authored-by: walker-ai <2398833647@qq.com> * Apply dsv3 router gemm kernel for deepseek-r1 fp4 (sgl-project#7677) * [AMD] Temporarily disable test_no_overlap_scheduler and test_vision_chunked_prefill (sgl-project#7717) * [RL] add --skip-warmup (sgl-project#7416) * [RL] support update_weights_from_distributed with different group and multiple weights (sgl-project#7292) * [router] add --log-level to sgl-router (sgl-project#6512) * [b200] support trt-llm allreduce fuse rms_norm_add kernel (sgl-project#7621) * [CPU] Bind threads and numa node for each TP rank (sgl-project#6549) Co-authored-by: srinarayan-srikanthan <srinarayan.srikanthan@intel.com> * Support non-contiguous query input for extend/decode attention (sgl-project#7462) * Support updating weights at once by stopping all requests (sgl-project#6698) Signed-off-by: Tianyu Zhou <albert.zty@antgroup.com> Co-authored-by: Zilin Zhu <zhuzilinallen@gmail.com> * Fix num_tokens_pre_allocated in disaggregation log (sgl-project#7714) * [CPU] [sgl-kernel] set dispatch key of initialize to CatchAll (sgl-project#7734) * [CPU] fix all_reduce and all_gather (sgl-project#6770) Co-authored-by: blzheng <beilei.zheng@intel.com> * fix awq and dsv3 fused gemm compatible (sgl-project#7735) * [CI][Router] Fix bench_one_batch_server for pd router test (sgl-project#7731) Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> * Add CUTLASS FP8 Blockscale MoE kernel for Hopper architecture (sgl-project#7278) Co-authored-by: HydraQYH <QYH820@Outlook.com> Co-authored-by: TianQiLin666666 <1834987979@qq.com> * fix dsv3 fused proj check (sgl-project#7738) * Ascend attention backend(PA&MLA) (sgl-project#7722) Co-authored-by: Maksim <makcum888e@mail.ru> Co-authored-by: VDV1985 <vladdv85@mail.ru> * [fix] fix dsv3_router_gemm filter (sgl-project#7750) * [CPU] refine CPU integration code (sgl-project#7647) * [CPU] support the case where num_attention_heads or intermediate_size is not divisible by the TP size (sgl-project#6771) * support qwen3 dense model dp attention (sgl-project#7681) * [optimize] add two stream norm for qwen3 (sgl-project#7740) Co-authored-by: ispobock <ispobaoke@gmail.com> * feat: use D2D instead of H2H in pp (sgl-project#7673) Co-authored-by: alpha-baby <fujianhao1997@qq.com> * [Bug] add flashinfer bool check for fusedmoe in Qwen moe models (sgl-project#7723) * [fix] put cpu in the first priority in get_device() (sgl-project#7752) * [optimize] fuse renormalize into moe_topk_softmax (sgl-project#7744) Co-authored-by: ispobock <ispobaoke@gmail.com> * chore: bump sgl-kernel 0.2.2 (sgl-project#7755) * fix CI: update native api ipynb (sgl-project#7754) Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> * fuse renormal into moe topk softmax kernel python code (sgl-project#7751) Co-authored-by: ispobock <ispobaoke@gmail.com> Co-authored-by: zhyncs <me@zhyncs.com> * Remove type conversion and fix id map in topk (sgl-project#7759) * Add V2-lite model test (sgl-project#7390) Co-authored-by: DiweiSun <105627594+DiweiSun@users.noreply.github.com> * refactor llama4 dp attention logic (sgl-project#7729) * fix(docs): fix the broken link in `docs/references/production_metrics.md` (sgl-project#7741) Signed-off-by: rudeigerc <rudeigerc@gmail.com> * [fix] update bench_speculative.py for compatibility (sgl-project#7764) Signed-off-by: Kay Yan <kay.yan@daocloud.io> * Move mem_fraction_static adjustment for multimodal models to `server_args.py` & Fix session control & Other cleanups (sgl-project#7748) * [RL] Add --nccl-port to prevent port conflict (sgl-project#7418) * [RL] add pause and continue generation for async rl training (sgl-project#7419) * [Fix] Alloc return type error (sgl-project#7778) Signed-off-by: Capronir <839972205@qq.com> * [feat] Support EAGLE3 for Qwen (sgl-project#7745) Co-authored-by: 纬杭 <ximing.wxm@antgroup.com> Co-authored-by: zyksir <zyksir@outlook.com> * saving hidden_states.clone() (sgl-project#7705) * [1/n]: add cutlass W4A8 moe kernel for hopper architecture (sgl-project#7772) Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com> * add model: qwen2-audio (sgl-project#7596) * Optimize Hopper CUTLASS FP8 Blockwise Grouped GEMM Kernel in Small K Scenario (sgl-project#7782) * Embedding parallel by attn_tp (sgl-project#7623) * fix: fix apply_shuffle_mul_sum (sgl-project#7444) * chore: bump sgl-kernel v0.2.3 (sgl-project#7784) * fix: use nvidia-nccl-cu12 2.27.5 (sgl-project#7787) * DP Attention with Auto DeepEP Dispatch (sgl-project#7222) * chore: upgrade sgl-kernel v0.2.3 (sgl-project#7786) * Fix incorrect spec_num_draft_tokens in draft_extend (sgl-project#7757) * [fix] fix misusing of is_cuda (sgl-project#7790) * Add treemask mode to build_eagle_tree & release sgl-kernel 0.2.3 (sgl-project#7756) Co-authored-by: Pranjal Shankhdhar <pranjal.ssh@gmail.com> * chore: bump sgl-kernel v0.2.4 (sgl-project#7800) * ci: fix port args (sgl-project#7792) * Fix CI test OOM issue. (sgl-project#7799) * chore: upgrade sgl-kernel v0.2.4 (sgl-project#7801) * chore: bump v0.4.9 (sgl-project#7802) * fix merge conflict issue * fix hpu attention nonetyep issue * fix alignment * fix alignment2 * Ci failure fixes * fix attention-backend choices --------- Signed-off-by: Xinyuan Tong <justinning0323@outlook.com> Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com> Signed-off-by: ch-tiger1 <xyz@ch-tech.ip-ddns.com> Signed-off-by: huanglong <huanglong@linux.alibaba.com> Signed-off-by: Ata Fatahi <immrata@gmail.com> Signed-off-by: keru <rukeyang@gmail.com> Signed-off-by: Tianyu Zhou <albert.zty@antgroup.com> Signed-off-by: rudeigerc <rudeigerc@gmail.com> Signed-off-by: Kay Yan <kay.yan@daocloud.io> Signed-off-by: Capronir <839972205@qq.com> Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Signed-off-by: Mohit Sinha <msinha@habana.ai> Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com> Co-authored-by: KavioYu <67678385+yukavio@users.noreply.github.com> Co-authored-by: kavioyu <kavioyu@tencent.com> Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com> Co-authored-by: yhyang201 <47235274+yhyang201@users.noreply.github.com> Co-authored-by: kk <43161300+kkHuang-amd@users.noreply.github.com> Co-authored-by: wunhuang <wunhuang@amd.com> Co-authored-by: DiweiSun <105627594+DiweiSun@users.noreply.github.com> Co-authored-by: u4lr451 <u4lr451@gmail.com> Co-authored-by: austindeng <austindeng@tencent.com> Co-authored-by: tianqilin.99 <tianqilin.99@bytedance.com> Co-authored-by: Qiaolin Yu <liin1211@outlook.com> Co-authored-by: ch-wan <cwan39@gatech.edu> Co-authored-by: Yijie Zhu <762412795@qq.com> Co-authored-by: 刁莹煜 <diaoyingyu1@hisilicon.com> Co-authored-by: Charles Chen <pychen96@gmail.com> Co-authored-by: Chang Su <chang.s.su@oracle.com> Co-authored-by: AniZpZ <zhuangsen.zp@antgroup.com> Co-authored-by: Yineng Zhang <me@zhyncs.com> Co-authored-by: shangmingc <caishangming@linux.alibaba.com> Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu> Co-authored-by: YanbingJiang <yanbing.jiang@intel.com> Co-authored-by: Wu, Chunyuan <chunyuan.wu@intel.com> Co-authored-by: jianan-gu <jianan.gu@intel.com> Co-authored-by: sdp <sdp@gnr799219.jf.intel.com> Co-authored-by: Binyao Jiang <byjiang1996@gmail.com> Co-authored-by: ishandhanani <82981111+ishandhanani@users.noreply.github.com> Co-authored-by: linzhuo <15313137931lz@gmail.com> Co-authored-by: ch-tiger1 <tiger@ch-tech.ip-ddns.com> Co-authored-by: ch-tiger1 <xyz@ch-tech.ip-ddns.com> Co-authored-by: fzyzcjy <5236035+fzyzcjy@users.noreply.github.com> Co-authored-by: ybyang <10629930+whybeyoung@users.noreply.github.com> Co-authored-by: Simo Lin <linsimo.mark@gmail.com> Co-authored-by: Jinn <47354855+jhinpan@users.noreply.github.com> Co-authored-by: Stefan He <hebiaobuaa@gmail.com> Co-authored-by: DarkSharpness <76582120+DarkSharpness@users.noreply.github.com> Co-authored-by: Atream <80757050+Atream@users.noreply.github.com> Co-authored-by: Li Hui <lambert80.ios@gmail.com> Co-authored-by: Huang Long <121648372+LLLL114@users.noreply.github.com> Co-authored-by: woodx <124784234+woodx9@users.noreply.github.com> Co-authored-by: Ata Fatahi <immrata@gmail.com> Co-authored-by: strgrb <zhangkaihong.zkh@antgroup.com> Co-authored-by: Zhang Kaihong <zhangkaihong.zkh@alibaba-inc.com> Co-authored-by: Wenbo Yang <solrex@users.noreply.github.com> Co-authored-by: Chang Su <csu272@usc.edu> Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com> Co-authored-by: Keyang Ru <rukeyang@gmail.com> Co-authored-by: ehuaa <ehuamail@163.com> Co-authored-by: pansicheng <sicheng.pan.chn@gmail.com> Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com> Co-authored-by: Jin Pan <jpan236@wisc.edu> Co-authored-by: Lifu Huang <lifu.hlf@gmail.com> Co-authored-by: Trevor Morris <tmorris@nvidia.com> Co-authored-by: JieXin Liang <Alcanderian@users.noreply.github.com> Co-authored-by: alcanderian <alcanderian@gmail.com> Co-authored-by: Ke Bao <ISPObaoke@163.com> Co-authored-by: Sai Enduri <saimanas.enduri@amd.com> Co-authored-by: Yi Zhang <1109276519@qq.com> Co-authored-by: xutizhou <xutingz@nvidia.com> Co-authored-by: TianQiLin666666 <1834987979@qq.com> Co-authored-by: HAI <hixiao@gmail.com> Co-authored-by: Yuhong Guo <guoyuhong1985@outlook.com> Co-authored-by: huangtingwei <141888744+huangtingwei9988@users.noreply.github.com> Co-authored-by: Alex Sun <alex.s@amd.com> Co-authored-by: valarLip <103567126+valarLip@users.noreply.github.com> Co-authored-by: Francis <38564764+ssssnow@users.noreply.github.com> Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: xianzhiT <xianzhitang@tencent.com> Co-authored-by: yilian49 <43861414+yilian49@users.noreply.github.com> Co-authored-by: DangKai <dangkai4u@outlook.com> Co-authored-by: dangkai.dk <dangkai.dk@alibaba-inc.com> Co-authored-by: Thien Tran <gau.nernst@yahoo.com.sg> Co-authored-by: ll819214 <18801269230@163.com> Co-authored-by: Li Junwen <lijunwen13@hisilicon.com> Co-authored-by: zixuanzhang226 <zixuanzhang@bytedance.com> Co-authored-by: Hongbo Xu <1320612015@qq.com> Co-authored-by: shangmingc <csmthu@gmail.com> Co-authored-by: eigen <52445717+yyihuang@users.noreply.github.com> Co-authored-by: mlmz <54172054+minleminzui@users.noreply.github.com> Co-authored-by: Ruihang Lai <ruihangl@cs.cmu.edu> Co-authored-by: Meng, Peng <pengmeng@tencent.com> Co-authored-by: Mick <mickjagger19@icloud.com> Co-authored-by: yhyang201 <yhyang201@gmail.com> Co-authored-by: tarinkk <129432511+tarinkk@users.noreply.github.com> Co-authored-by: tarinkk <rt572@physics.rutger.edu> Co-authored-by: tarinkk <rt572@rutgers.physics.edu> Co-authored-by: Hanming Lu <69857889+hanming-lu@users.noreply.github.com> Co-authored-by: Zheng, Beilei <beilei.zheng@intel.com> Co-authored-by: Sheng Qi <shengqi2018@pku.edu.cn> Co-authored-by: finetune <82650881+finetunej@users.noreply.github.com> Co-authored-by: Hubert Lu <55214931+hubertlu-tw@users.noreply.github.com> Co-authored-by: Kan Wu <wukanustc@gmail.com> Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com> Co-authored-by: narutolhy <582909902@qq.com> Co-authored-by: lukec <118525388+sleepcoo@users.noreply.github.com> Co-authored-by: shuaills <shishuaiuoe@gmail.com> Co-authored-by: Shenggui Li <somerlee.9@gmail.com> Co-authored-by: Yingyi Huang <yingyihuang2000@outlook.com> Co-authored-by: Simon_CQK <cqk0100@gmail.com> Co-authored-by: Kyungmin Lee <30465912+lkm2835@users.noreply.github.com> Co-authored-by: 晟海 <huangtingwei.htw@antgroup.com> Co-authored-by: yych0745 <1398089567@qq.com> Co-authored-by: HandH1998 <1335248067@qq.com> Co-authored-by: 弋云 <yiyun.wyt@antgroup.com> Co-authored-by: walker-ai <2398833647@qq.com> Co-authored-by: Zilin Zhu <zhuzilinallen@gmail.com> Co-authored-by: srinarayan-srikanthan <srinarayan.srikanthan@intel.com> Co-authored-by: Albert <albert.zty@antgroup.com> Co-authored-by: Ziming Huang <1520787127@qq.com> Co-authored-by: ayrnb <70835312+ayrnb@users.noreply.github.com> Co-authored-by: HydraQYH <QYH820@Outlook.com> Co-authored-by: ronnie_zheng <zl19940307@163.com> Co-authored-by: Maksim <makcum888e@mail.ru> Co-authored-by: VDV1985 <vladdv85@mail.ru> Co-authored-by: ispobock <ispobaoke@gmail.com> Co-authored-by: TianyuZhang1214 <tianyuzhang1214@163.com> Co-authored-by: alpha-baby <fujianhao1997@qq.com> Co-authored-by: Yuchen Cheng <rudeigerc@gmail.com> Co-authored-by: Kay Yan <kay.yan@daocloud.io> Co-authored-by: Caproni <40862361+Capronir@users.noreply.github.com> Co-authored-by: Ximingwang-09 <72070413+Ximingwang-09@users.noreply.github.com> Co-authored-by: 纬杭 <ximing.wxm@antgroup.com> Co-authored-by: zyksir <zyksir@outlook.com> Co-authored-by: SijiaYang <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com> Co-authored-by: Leng Yue <lengyue@lengyue.me> Co-authored-by: Qi Yuhang <45795032+HydraQYH@users.noreply.github.com> Co-authored-by: Gang Chen <13298548+MoonBall@users.noreply.github.com> Co-authored-by: Pranjal Shankhdhar <pranjal.ssh@gmail.com> Co-authored-by: jay <jthakur@habana.ai>

…ct#7772) Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com>

feat: add cutlass W4A8 MoE kernel for Hopper architecture

9a951e2

Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com> Co-authored-by: yicwang <yichen.wang@bytedance.com>

yangsijia-serena requested review from zhyncs, ispobock, HandH1998, BBuf, yizhang2077, merrymercy, FlamingoPg and HaiShaw as code owners July 4, 2025 08:06

gemini-code-assist bot reviewed Jul 4, 2025

View reviewed changes

AniZpZ self-requested a review July 4, 2025 08:11

AniZpZ reviewed Jul 4, 2025

View reviewed changes

yangsijia-serena added 2 commits July 4, 2025 23:15

fix: set non-zero default value for the parameters of cutlass_w4a8_mo…

89bbb6e

…e_mm. Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com>

ut: add multi grouped gemm ut in test_cutlass_w4a8_moe_mm.py

d5109aa

Signed-off-by: yangsijia.614 <yangsijia.614@bytedance.com>

zhyncs assigned AniZpZ, TianQiLin666666, HydraQYH, yue-zhang-2025, Alcanderian and BBuf Jul 5, 2025

AniZpZ approved these changes Jul 5, 2025

View reviewed changes

zhyncs merged commit da3890e into sgl-project:main Jul 5, 2025
35 of 39 checks passed

yangsijia-serena mentioned this pull request Jul 6, 2025

feat: support DeepSeek-R1-W4AFP8 model with ep-moe mode #7762

Merged

6 tasks

AniZpZ mentioned this pull request Jul 15, 2025

[1/n] chore: decouple quantization implementation from vLLM dependency #7992

Merged

6 tasks

yuhyao mentioned this pull request Aug 20, 2025

[Bug] Fix w4afp8 moe kernel #9392

Merged

4 tasks

pandengyao mentioned this pull request Aug 26, 2025

[feat] DeepSeek-R1 W4AFP8: unify ep_moe and support DeepEP normal & low_latency #9598

Open

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[1/n]: add cutlass W4A8 moe kernel for hopper architecture #7772

[1/n]: add cutlass W4A8 moe kernel for hopper architecture #7772

Uh oh!

yangsijia-serena commented Jul 4, 2025

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

AniZpZ Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

gemini-code-assist bot Jul 4, 2025

Uh oh!

AniZpZ Jul 4, 2025

Uh oh!

AniZpZ Jul 4, 2025 •

edited

Loading

Uh oh!

yangsijia-serena Jul 4, 2025

Uh oh!

Uh oh!

Uh oh!

	#define GROUP_SIZE 128
	static constexpr int GROUP_SIZE = 128;


		#include <iostream>

		constexpr uint64_t THREADS_PER_EXPERT = 512;

[1/n]: add cutlass W4A8 moe kernel for hopper architecture #7772

[1/n]: add cutlass W4A8 moe kernel for hopper architecture #7772

Uh oh!

Conversation

yangsijia-serena commented Jul 4, 2025

Motivation

Modifications

Checklist

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

AniZpZ Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

AniZpZ Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

AniZpZ Jul 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

yangsijia-serena Jul 4, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

AniZpZ Jul 4, 2025 •

edited

Loading