COMMITS
July 22, 2026
July 21, 2026
E
docs: Update README with new image tags
Eta committed
E
feat: Update to CUDA 12.8.2 and 12.9.2
Eta committed
E
feat: Update to CUDA 13.0.3 and 13.1.2
Eta committed
E
docs: Update README with new image tags
Eta committed
E
feat: Update to CUDA 13.3.0
Eta committed
E
feat: Update NCCL to 2.30.7-1 and HPC-X to 2.50
Eta committed
July 8, 2026
A
nccl test alignment with backend sweep for manual execution/debug.
Aleksandr Malyshev committed
A
chore: Remove commented command for interactive testing in nccl-test-distributed-8x8-ib.yaml
Aleksandr Malyshev committed
June 26, 2026
June 24, 2026
A
chore: rename nccl-test-distributed-h100.yaml template file as it applicable for all 8x8 ib configs.
Aleksandr Malyshev committed
A
Merge branch 'master' into amalyshev/ncc_test_alignment_with_backend_sweep
Aleksandr Malyshev committed
A
chore: Enhance run_test.py with additional benchmark templates and support for nonvlink fabric option
Aleksandr Malyshev committed
June 23, 2026
E
Merge pull request #95 from coreweave/eta/doca-ofed
Eta committed
June 18, 2026
A
chore: Update run_test.py to support new platform and fabric options, and enhance benchmark command templates
Aleksandr Malyshev committed
E
docs: Update README with new image tags
Eta committed
A
chore: Update NCCL test image version to 13.2.1 for Ubuntu 24.04 in nccl-test-distributed-h100.yaml
Aleksandr Malyshev committed
A
chore: Update NCCL test image version to 13.2.1 in nccl-test-distributed-h100.yaml
Aleksandr Malyshev committed
June 17, 2026
A
chore: Remove Mellanox OFED installation from Dockerfile and add run/wait scripts for NCCL tests
Aleksandr Malyshev committed
A
chore: Reintroduce Mellanox OFED installation in Dockerfile and update RDMA library installation process
Aleksandr Malyshev committed
June 15, 2026
A
Merge branch 'master' into amalyshev/ncc_test_alignment_with_backend_sweep
Aleksandr Malyshev committed
A
feat: Enhance environment variable handling in run_test.py for MPI jobs
Aleksandr Malyshev committed
June 11, 2026
A
chore: Downgrade CUDA version to 13.2.1 and update NCCL tests commit in Dockerfile and workflow configurations
Aleksandr Malyshev committed
A
chore: Update CUDA version to 13.2.75 and adjust NCCL test configurations in YAML files
Aleksandr Malyshev committed
A
chore: Remove Mellanox OFED installation from Dockerfile and update comments regarding RDMA libraries
Aleksandr Malyshev committed
A
chore: Update CUDA version to 13.2.75 and NCCL image tags to v2.30.4-1 in configuration files
Aleksandr Malyshev committed
W
feat: bump NCCL test images to cuda 13.2.1 / nccl 2.30.4
Willy Markuske committed
June 9, 2026
A
feat: Add run and wait scripts for NCCL tests
Aleksandr Malyshev committed