Skip to content

Issues: NVIDIA/nccl

Beta
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

743 Open 739 Closed
743 Open 739 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

NCCL all2all test Busbw drops after PXN is enabled
#1697 opened Apr 24, 2025 by gatesking
Segmentation fault (core dumped) when running nccl-tests bug more info needed
#1696 opened Apr 23, 2025 by misslxs
1
enhancement of NCCL_PARAM enhancement
#1695 opened Apr 22, 2025 by GeofferyGeng
1
Question about NCCL communicate process and HCA
#1694 opened Apr 21, 2025 by marks221b
NCCL Multi-node Topology question triaged
#1693 opened Apr 20, 2025 by ramgandikota
1
Question about NCCL2.27 sym memory usage in Cuda Graph
#1692 opened Apr 20, 2025 by zyksir
1
Question about the P2P comm warmup process
#1691 opened Apr 19, 2025 by Edenzzzz
3
enable inplace ring-all_reduce with userbuff
#1689 opened Apr 18, 2025 by visualxu
question about GDR Flush Operation
#1688 opened Apr 18, 2025 by fanfanfanWu
1
Question about adaptive routing question
#1687 opened Apr 17, 2025 by GeofferyGeng
2
All-to-all seems to trigger connection of many channels thus a large memory usage? enhancement
#1685 opened Apr 16, 2025 by kwen2501
3
failed to IPC register userbuff more info needed question
#1684 opened Apr 16, 2025 by youzhedian
2
Misplaced const qualifiers in nccl.h enhancement
#1681 opened Apr 14, 2025 by blazej-smorawski
1
How to set griddim for ncclsend/recv? question
#1680 opened Apr 14, 2025 by chenhongyu2048
What is the right way to capture ncclAllreduce only? question
#1679 opened Apr 12, 2025 by ghostplant
Why is there a significant performance difference between using the same ring and two reverse rings for Broadcast command question
#1678 opened Apr 11, 2025 by hennry205
How to set a timeout for ncclCommInitRank? question triaged
#1677 opened Apr 10, 2025 by hiworldwzj
1
Recent DGX 25.01 cause huge performance regression for NCCL <= 2.22.3 enhancement question
#1676 opened Apr 8, 2025 by xman1979
3
using avx to accelerate inter node protoll enhancement
#1675 opened Apr 7, 2025 by visualxu
nvidia kvm pcie passthrough how topo is configured for virtual machines more info needed question
#1674 opened Apr 7, 2025 by Vagony
3
transport/nvls.cc:598 NCCL WARN Cuda failure 1 'invalid argument' more info needed question
#1672 opened Apr 6, 2025 by cassanof
1
Is it possible to also avoid cross rail when channel count is odd? enhancement question
#1671 opened Apr 5, 2025 by fishautumn
Identifying the GPU to NIC mapping used by NCCL via nccl logs question triaged
#1667 opened Apr 2, 2025 by pradeep-hk
7
Unexpected delay in Concurrent Groups with Different Rank Orders question triaged
#1666 opened Apr 1, 2025 by amjal
1
Kind Query about Why NCCL produce ".o" file for each datatype, instead of pass hyperparameters, e.g. "Datatype t" question
#1665 opened Apr 1, 2025 by CVPR-2020
ProTip! Follow long discussions with comments:>50.