3 people 2b85d9cb65 [Perf][Bugfix] Prefer MNNVL for single-node FlashInfer allreduce again (#61989) ...

Signed-off-by: khluu <khluu000@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Wei Zhao <51183510+wzhao18@users.noreply.github.com>

2026-10-17 02:55:28 -07:00

..

__init__.py

[Core][Refactor] move parallel_utils into vllm/distributed (#3950)

2024-04-10 15:33:30 -07:00

aiter_custom_all_reduce.py

[ROCm] Enable the cuMem CUDA-graph pool offload on ROCm (#59523)

2026-10-09 08:29:55 +00:00

all2all.py

[ROCm][DSv4] Fuse the shared expert into AITER MegaMoE (#61004)

2026-10-12 15:53:16 +08:00

all_reduce_utils.py

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

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

base_device_communicator.py

[Perf][MoE] Eliminate staging from NCCL symmetric reduce-scatter (#49194)

2026-10-06 07:37:01 -07:00

cpu_communicator.py

Add pydocstyle to the ruff rules (#52136)

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

cuda_communicator.py

[Feature] Support batch invariance on ROCm (#52231)

2026-10-12 10:53:26 -05:00

cuda_wrapper.py

[Bugfix][KV Offload] Register the offload region in chunks (#51081)

2026-09-17 13:19:56 +03:00

custom_all_reduce.py

[Feature] Release the CUDA graph pool on sleep (#59160)

2026-10-04 01:13:10 +00:00

flashinfer_all_reduce.py

[Perf][Bugfix] Prefer MNNVL for single-node FlashInfer allreduce again (#61989)

2026-10-17 02:55:28 -07:00

flashinfer_pcie_ipc_all_reduce.py

[Distributed] Add opt-in FlashInfer PCIe IPC all-reduce backend (#53576)

2026-09-01 03:34:56 +00:00

mnnvl_compat.py

[Kernel] Add FlashInfer MoE A2A Kernel (#36022)

2026-03-15 23:45:32 -07:00

pynccl_allocator.py

[Core] Model console logging as CLI configuration (#57205)

2026-09-25 11:36:59 +01:00

pynccl_wrapper.py

[Distributed] Support pre-shared ncclUniqueId rendezvous for weight transfer (#53784)

2026-09-01 14:55:57 +08:00

pynccl.py

Add pydocstyle to the ruff rules (#52136)

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

quick_all_reduce.py

[Misc] Consolidate csrc shared headers and prune dead code (#60721)

2026-10-12 09:46:22 -07:00

shm_broadcast.py

[Bugfix][Distributed] Own SHM out-of-band buffers before releasing read slots (#61000)

2026-10-12 01:44:12 -04:00

shm_object_storage.py

[Bugfix][Multimodal] Fix reference counting of the SHM processor cache (#59695)

2026-10-02 22:14:13 +08:00

shm_tensor_arena.py

[Core] shm tensor arena for efficient cpu->gpu worker broadcast (#51207)

2026-10-09 16:13:24 -07:00

symm_mem.py

[Rubin] Enable NVLink all-reduce paths on SM107 (#49647)

2026-07-29 20:53:25 +00:00

xpu_communicator.py

[Feat][XPU] VLLM_BATCH_INVARIANT support for Dense/MoE models (#55881)

2026-09-22 11:06:54 -04:00