Skip to content

Issues: openucx/ucx

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

601 Open 1,483 Closed
601 Open 1,483 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

Intra-node communication fails when using Nvidia hpc-x in a Singularity container Bug
#10404 opened Jan 6, 2025 by mredenti
test/apps/test_cuda_get_symbol_address.cu build failure on Debian 12 with "relocation R_X86_64_32S against `.rodata' can not be used when making a PIE object; recompile with -fPIE" Bug
#10402 opened Jan 4, 2025 by hexclover
@rakhmets
1
When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked Bug
#10370 opened Dec 12, 2024 by ivanallen
15
How to create a group communication with UCX?
#10361 opened Dec 7, 2024 by MoFHeka
2
UCT/CUDA_IPC: Possible UB when enabling CUDA_IPC_CACHE
#10346 opened Dec 2, 2024 by andylin-hao
Shared memory transport provides lower throughput for large message sizes than inter-node transport via the Infiniband Bug
#10317 opened Nov 21, 2024 by satishskamath
OpenMPI+UCX with multiple GPUs error: "named symbol not found" Bug
#10304 opened Nov 15, 2024 by pascal-boeschoten-hapteon
3
How to initialize UCP when there are multiple GPUs in one machine and multiple GPU machines in cluster?
#10276 opened Nov 5, 2024 by MoFHeka
CUDA-Aware UCX with a mixture of CPU-only & GPU Nodes Bug
#10273 opened Nov 4, 2024 by judicaelclair
Single network card is very fast, dual network card speed is super slow Bug
#10249 opened Oct 23, 2024 by yangrudan
3
Failed to run ucx assessment job and getting job limit exceed 12 while triggering ucx assessment job [version ucx 0.38 version onwards,facing this issue-unable to trigger ucx assessment job] Bug
#10248 opened Oct 22, 2024 by pradeep5561
1
Q: UCX support for Gracehopper + Slingshot 11 Bug
#10234 opened Oct 16, 2024 by angainor
Can a chain be built between ucx 1.12 and ucx1.14
#10222 opened Oct 11, 2024 by super-train
2
Incorrect result report for UCP tag_bw test Bug
#10184 opened Sep 26, 2024 by SeyedMir
@yosefe
How preallocate buffer through rendezvous protocol before ucp_tag_recv_nbx actually receiving?
#10148 opened Sep 12, 2024 by MoFHeka
3
Is there any benchmark of P2P communication between NCCL and UCX(ucp)?
#10147 opened Sep 11, 2024 by MoFHeka
rocm_ipc_md.c:79 UCX ERROR Failed to create ipc for 0x7f9030c18000/8000 Bug
#10087 opened Aug 25, 2024 by jinz2014
@edgargabriel
14
Address Registration Error in CUDA Aware MPICH 4.2.2 + UCX 1.17.0 Application
#10085 opened Aug 23, 2024 by cl3to
@amastbaum
9
When testing ROCm D2D transfers with UCX_TLS=rc, how does setting UCX_IB_GPU_DIRECT_RDMA=0 affect the osu_bw test results?
#10077 opened Aug 20, 2024 by shuiYizero
5
how does use PCIe peer-to-peer or NVLink between two containers that each have an isolated GPU
#10070 opened Aug 20, 2024 by linxiaochou
3
UCX ignores exclusively setting TCP devices when RoCE is available. Bug
#10049 opened Aug 6, 2024 by bertiethorpe
8
Assertion `worker->inprogress++ == 0' failed Bug
#10039 opened Aug 2, 2024 by pereverges
3
How to change single copy VIA xpmem execution to the sender process
#10019 opened Jul 17, 2024 by arun-chandran-edarath
3
How to Use Shared Memory for Intra-node Inter-process Data Transfer? Bug
#10016 opened Jul 16, 2024 by Clownier
8
Stuck at waitForEvents Bug
#10015 opened Jul 16, 2024 by pereverges
1
ProTip! Mix and match filters to narrow down what you’re looking for.