uccl-project / uccl
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
Why this job: matched by what the repo talks about: gpu, cache, weight, efficient
What it does
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., IBGDA), with two key focuses:
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit yesterday
- Stars
- 1,582
- +9 this week
- Licence
- Apache-2.0
- Allows business use
- Downloads for
- Linux
- From the latest release
- Latest version
- v0.1.1
- Released 6 months ago
- Built with
- C++
- 58% of the code
Is it looked after?
- Contributors
- 20
- Last 90 days
- Issues
- 79% closed
- 52 still open
- Pull requests
- 797 merged
- 14 waiting
- Releases
- 2
- 2 in the last year
- No contributing guide
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| kvcache-ai / Mooncake | 7,025 | today | Apache-2.0 | Linux |
| flashinfer-ai / flashinfer | 6,900 | today | Apache-2.0 | Linux |
| vllm-project / vllm | 96,774 | today | Apache-2.0 | Mac, Linux |
More: download files, languages, topics and project record
Languages
- C++ 58%
- Cuda 16%
- Python 15%
- Jupyter Notebook 3%
Topics
- ai
- allreduce
- amd
- broadcom
- collective
- cuda
- gpu
- hpc
- llm
- networking
- nvidia
- rdma
Project record
- JobLocal LLMs
- StatusListed
- Commits1,656
- Forks198
- Watching31
- Stars in 28 days+41
- CreatedJan 6, 2025
- First seen by usNov 7, 2026
- CheckedNov 12, 2026