GitHub · C++
NVIDIA/nccl
Optimized primitives for collective multi-GPU communication
- Stars
- 5,250
- Forks
- 1,502
- Stars gained in 24 hours
- +47
- Forks gained in 24 hours
- +38
Topics: communications, cpp, cuda, deep-learning, gpu, nvidia
Latest data point: Nov 22, 2026 · 5,250 stars · 1,502 forks