Commits · 17edd8a807019c8d1e58634aecb1de7984e8d467 · OpenDAS / vllm_cscc

04 Oct, 2025 1 commit

[cpu][perf] Accelerate unquantized-linear for AArch64 through oneDNN/ACL and... · 9705fba7

Fadi Arafeh authored Oct 04, 2025


[cpu][perf] Accelerate unquantized-linear for AArch64 through oneDNN/ACL and weight prepack (#25948)
Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>
Co-authored-by: Li, Jiang <jiang1.li@intel.com>

9705fba7

03 Oct, 2025 5 commits
- Avoid division by zero in cache DS MLA kernel (#26174) · 300a59c4
  Matthew Bonanni authored Oct 03, 2025
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  300a59c4
- Fix undefined symbol: cutlass_moe_mm_sm100 (#26098) · 4f8f47e8
  Jun Jiang authored Oct 03, 2025
```
Signed-off-by: Jun Jiang <jasl9187@hotmail.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
```
  4f8f47e8
- [Perf] Optimize `reshape_and_cache` CUDA Kernel (#25955) · eb0fa438
  Jiangyun Zhu authored Oct 03, 2025
```
Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
Co-authored-by: Liu-congo <1502632128@qq.com>
```
  eb0fa438
- [DeepSeek] Improve performance of DS MLA cache kernel (#26132) · 47b93395
  Matthew Bonanni authored Oct 02, 2025
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  47b93395
- [CI/Build] Conditionally register cutlass_fp4_group_mm to fix building on Hopper (#26138) · 5d5146ee
  Michael Goin authored Oct 02, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  5d5146ee
01 Oct, 2025 6 commits

[ROCm][Build] Add support for AMD Ryzen AI MAX / AI 300 Series (#25908) · aac622e0
Hosang authored Oct 01, 2025
```
Signed-off-by: Hosang Yoon <hosang.yoon@amd.com>
```
aac622e0

[BugFix][DP/EP] Fix CUTLASS MLA hang under load (#26026) · 1726e93e

Lucas Wilkinson authored Oct 01, 2025


Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>

1726e93e

[NVIDIA] Blackwell Family (#24673) · 5234dc74

Johnny authored Oct 01, 2025


Signed-off-by: Johnny <johnnynuca14@gmail.com>
Signed-off-by: johnnynunez <johnnynuca14@gmail.com>
Signed-off-by: Johnny <johnnync13@gmail.com>
Signed-off-by: Salvatore Cena <cena@cenas.it>
Co-authored-by: Aidyn-A <31858918+Aidyn-A@users.noreply.github.com>
Co-authored-by: Salvatore Cena <cena@cenas.it>

5234dc74

[Bugfix] Fix `__syncwarp` on ROCM (#25996) · e952eee6
Zhewen Li authored Sep 30, 2025

e952eee6
Fix INT8 quantization error on Blackwell GPUs (SM100+) (#25935) · 99028fda
Param authored Sep 30, 2025
```
Signed-off-by: padg9912 <phone.and.desktop@gmail.com>
```
99028fda

Update launch_bounds_utils.h for correct compile on Multiple Cuda Arch - PTXAS... · a73f6491

Salvatore Cena authored Oct 01, 2025

Update launch_bounds_utils.h for correct compile on Multiple Cuda Arch - PTXAS out of range Warning (#25843)
Signed-off-by: Salvatore Cena <cena@cenas.it>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

a73f6491

30 Sep, 2025 1 commit

[New Model] DeepSeek-V3.2 (Rebased to Main) (#25896) · fa7e254a

Yongye Zhu authored Sep 30, 2025


Signed-off-by: Chen Zhang <zhangch99@outlook.com>
Signed-off-by: youkaichao <youkaichao@gmail.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: NickLucche <nlucches@redhat.com>
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
Signed-off-by: Barry Kang <43644113+Barry-Delaney@users.noreply.github.com>
Signed-off-by: Lucia Fang <fanglu@meta.com>
Co-authored-by: Chen Zhang <zhangch99@outlook.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: yewentao256 <zhyanwentao@126.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Lucia Fang <116399278+luccafong@users.noreply.github.com>
Co-authored-by: Lucia Fang <fanglu@meta.com>
Co-authored-by: NickLucche <nlucches@redhat.com>
Co-authored-by: Siyuan Fu <siyuanf@nvidia.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Xiaozhu Meng <mxz297@gmail.com>
Co-authored-by: Barry Kang <43644113+Barry-Delaney@users.noreply.github.com>

fa7e254a

26 Sep, 2025 1 commit
- Kernel-override Determinism [1/n] (#25603) · dc48ba0c
  Bram Wasti authored Sep 26, 2025
```
Signed-off-by: Bram Wasti <bwasti@meta.com>
```
  dc48ba0c
25 Sep, 2025 2 commits

[Hardware][RISC-V] Add riscv64 support for vLLM with scalar (#22112) · 1e9a77e0
chenlang authored Sep 25, 2025
```
Signed-off-by: chenlang <chen.lang5@zte.com.cn>
Co-authored-by: chenlang <10346245@zte.com.cn>
```
1e9a77e0

[Model] Add LongCat-Flash (#23991) · 845adb3e

XuruiYang authored Sep 25, 2025


Signed-off-by: yangxurui <yangxurui@meituan.com>
Co-authored-by: yangxurui <yangxurui@meituan.com>

845adb3e

24 Sep, 2025 3 commits
- [MISC] replace c10::optional with std::optional (#25602) · 5c1e496a
  Shiyan Deng authored Sep 24, 2025
```
Signed-off-by: Shiyan Deng <dsy842974287@meta.com>
```
  5c1e496a
- [fix]: add Arm 4bit fused moe support (#23809) · 359d2930
  Nikhil Gupta authored Sep 24, 2025
```
Signed-off-by: Nikhil Gupta <nikhil.gupta2@arm.com>
```
  359d2930
- [Compile] Fix AMD Compile Error (#25518) · bf68fd76
  Wentao Ye authored Sep 23, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  bf68fd76
23 Sep, 2025 4 commits

[Bugfix] [B200] cutlass_mla - ensure kv_split == 1 for batch size > 1 (#25509) · 1210e4d9
Alexander Matveev authored Sep 23, 2025
```
Signed-off-by: Alexander Matveev <amatveev@redhat.com>
```
1210e4d9

[ROCm] Add skinny gemm bias support for dtypes fp16,bf16,fp8 (#24988) · a3a78280

Hashem Hashemi authored Sep 23, 2025


Signed-off-by: Hashem Hashemi <hashem.hashemi@amd.com>
Signed-off-by: Hashem Hashemi <159079214+amd-hhashemi@users.noreply.github.com>

a3a78280

Use macro guard CUDA functions for back compatibility in grouped_topk_kernel.cu (#25346) · 527821d1

Ming Yang authored Sep 23, 2025


Signed-off-by: Ming Yang <minos.future@gmail.com>
Signed-off-by: Rahul Tuli <rtuli@redhat.com>
Co-authored-by: Rahul Tuli <rtuli@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Co-authored-by: Lu Fang <30275821+houseroad@users.noreply.github.com>
Co-authored-by: Ye (Charlotte) Qi <yeq@meta.com>

527821d1

[BugFix] Fix UB in per_token_group_quant.cu (#24913) · 2357480b
rivos-shreeasish authored Sep 23, 2025
```
Signed-off-by: Shreeasish Kumar <shreeasish@rivosinc.com>
```
2357480b

19 Sep, 2025 1 commit
- [Compile] Fix Compile Warning for Ignoring `MIN_BLOCK_PER_SM` (#25193) · 711e9129
  Wentao Ye authored Sep 19, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  711e9129
18 Sep, 2025 3 commits
- silu-v1: Fix EPS not being used during max-reduction (#25069) · abdfcd4f
  Elvir Crnčević authored Sep 18, 2025
```
Signed-off-by: elvircrn <elvircrn@gmail.com>
```
  abdfcd4f
- Fix: Add explicit #include <omp.h> for OpenMP compatibility on certain toolchains (#24951) · 4f02b77d
  ihb2032 authored Sep 18, 2025
```
Signed-off-by: lyd1992 <liuyudong@iscas.ac.cn>
Signed-off-by: ihb2032 <1355790728@qq.com>
```
  4f02b77d
- [Kernel] Better inf handling for grouped topk cu (#24886) · 81b16a2b
  Lumina authored Sep 18, 2025
```
Signed-off-by: lumina37 <starry.qvq@gmail.com>
```
  81b16a2b
17 Sep, 2025 6 commits
- [Bugfix] Fix accuracy issue for silu_mul + nvfp4 quant fusion kernel (#24833) · e6585ddb
  elvischenv authored Sep 18, 2025
```
Signed-off-by: elvischenv <219235043+elvischenv@users.noreply.github.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
```
  e6585ddb
- [Bugfix][B200] Fix `cutlass_mla` hang (#24966) · fedb75fa
  Alexander Matveev authored Sep 17, 2025
```
Signed-off-by: Alexander Matveev <amatveev@redhat.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
```
  fedb75fa
- [Kernel] Faster pre-processing time for W4A8 (#23972) · 3c068c63
  czhu-cohere authored Sep 17, 2025
```
Signed-off-by: czhu-cohere <conway.zhu@cohere.com>
```
  3c068c63
- Remove old cutlass mla (#23961) · 8f3616f4
  Matthew Bonanni authored Sep 17, 2025
```
Signed-off-by: Matthew Bonanni <mbonanni001@gmail.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  8f3616f4
- Apply fixes for CUDA 13 (#24599) · bfe93801
  Aidyn-A authored Sep 17, 2025
```
Signed-off-by: Aidyn-A <aidyn.b.aitzhan@gmail.com>
```
  bfe93801
- [Bugfix] Fix Stream usage in CPU model runner and OneDNN kernel check (#25046) · 9fccd04e
  Li, Jiang authored Sep 17, 2025
```
Signed-off-by: jiang1.li <jiang1.li@intel.com>
```
  9fccd04e
15 Sep, 2025 3 commits

[Bug] Fix Cutlass Scaled MM Compilation Error (#24887) · e757a629
Wentao Ye authored Sep 15, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
e757a629
[Transform] Deterministic Hadacore Transforms (#24106) · a0b26701
Kyle Sayers authored Sep 15, 2025
```
Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
```
a0b26701

Fp8 paged attention update (#22222) · 01413e0c

xiao-llm authored Sep 15, 2025


Signed-off-by: Xiao Yu <xiao.yu@amd.com>
Signed-off-by: xiao-llm <xiao.yu.dc@outlook.com>
Co-authored-by: Xiao Yu <xiao.yu@metamaterial.com>
Co-authored-by: Xiao Yu <xiao.yu@amd.com>
Co-authored-by: Bowen Bao <bowenbao@amd.com>

01413e0c

13 Sep, 2025 4 commits
- [CI Failure] Fix test_flashinfer_cutlass_mxfp4_mxfp8_fused_moe (#24750) · 59d7ffc1
  Michael Goin authored Sep 13, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  59d7ffc1
- [Kernels][DP/EP] Optimize Silu Kernel for R1 (#24054) · 98229db2
  Elvir Crnčević authored Sep 13, 2025
```
Signed-off-by: elvircrn <elvircrn@gmail.com>
```
  98229db2
- [Perf] Use NVIDIA hardware-accelerated instruction for float to fp8_e4m3 quantization (#24757) · dbeee384
  elvischenv authored Sep 13, 2025
```
Signed-off-by: elvischenv <219235043+elvischenv@users.noreply.github.com>
```
  dbeee384
- [Chore] Remove unused batched RoPE op & kernel (#24789) · 5febdc87
  Woosuk Kwon authored Sep 13, 2025
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  5febdc87