Issues
- 2
MPI-style INPLACE operations
#683 opened by vspetrov - 0
Some puzzled problems about algorithm
#1018 opened by Luckyton - 0
Some puzzled problems about algorithms
#1017 opened by Luckyton - 3
Do TL/nccl and rccl only support allgatherv?
#1014 opened by a-c-dream - 2
- 2
Build Errors with CL/DOCA_UROM
#1002 opened by wfaderhold21 - 0
- 1
- 0
add new alltoallv alg field
#983 opened by pangjuan8848 - 3
- 1
[RFE] EC/CUDA: mode to avoid using persistent kernel
#673 opened by marsaev - 7
segfault when nranks>=18
#981 opened by angainor - 5
Building with ROCm/HIP fails on a system without GPU
#969 opened by lahwaacz - 22
- 0
Configurable OOB/service collectives
#916 opened by miharulidze - 2
API/ABI compatibility of UCC
#968 opened by jakirkham - 2
1.3.0 release tag
#965 opened by zzzoom - 1
Fortran OpenMP+ target offload: TL_CUDA DEBUG cannot create CUDA TL context without active CUDA context
#952 opened by burlen - 0
Create a conda-forge package for the UCC library.
#950 opened by sandeepd-nv - 2
Include #847 in v1.3.0 RC
#947 opened by sandeepd-nv - 1
- 6
Issues regarding ucc_coll_args::tag
#893 opened by nirandaperera - 5
- 1
Minor build issue at ucc_ep_map_is_identity()
#892 opened by alex--m - 4
ucc_mem_map limitations
#883 opened by nirandaperera - 1
`ucc_context_create` seems to take a constant time
#881 opened by nirandaperera - 5
Provide a NCCL-like initialization mechanism
#705 opened by almogsegal - 2
Segfault when running `ucc_info`
#868 opened by nirandaperera - 2
- 1
sm_52 compilation error, test for cuda version > 11.6
#865 opened by torehl - 3
Version 1.2.0 Release (April 30th)
#714 opened by manjugv - 1
ucc_lib_get_attr doesn't produce output
#796 opened by marsaev - 6
make install failed
#853 opened by pangjuan8848 - 3
Reduce-Scatter in TL/SHARP
#840 opened by jywangx - 2
[Question] Relationship between CLs and TLs
#817 opened by anxietymonger - 1
- 4
Updating ucc.conf usages in Wiki->FAQ
#726 opened by shimmybalsam - 5
- 1
`gatherv` implementation
#603 opened by nirandaperera - 0
UCC v1.2 Release
#648 opened by manjugv - 1
Use collective active_set to send data to self
#674 opened by marsaev - 2
scatterv implementation
#665 opened by almogsegal - 2
UCC Spec v1.1
#640 opened by manjugv - 0
[RFE] EC/CUDA: using cuStreamMemOps
#672 opened by marsaev - 0
Hang in MPI_tests for triggered reduce operations
#668 opened by samnordmann - 1
Hang in TL/CUDA for triggered reduce_scatterv
#638 opened by samnordmann - 1
Potential memleak in TL/CUDA
#641 opened by vspetrov - 5
Request specific TLs on Team creation
#651 opened by almogsegal - 4
- 0