Skip to content

NCCL

The NVIDIA Collective Communications Library (NCCL) implements multi-GPU and multi-node collective communication primitives that are performance optimized for NVIDIA GPUs.

homepage: https://developer.nvidia.com/nccl

version versionsuffix toolchain
2.1.4 -CUDA-9.0.176 system
2.10.3 -CUDA-11.3.1 GCCcore/10.3.0
2.10.3 -CUDA-11.4.1 GCCcore/11.2.0
2.10.3 -CUDA-11.5.2 GCCcore/11.2.0
2.11.4 gcccuda/2019b
2.12.12 -CUDA-11.7.0 GCCcore/11.3.0
2.16.2 -CUDA-11.7.0 GCCcore/12.2.0
2.16.2 -CUDA-12.0.0 GCCcore/12.2.0
2.18.3 -CUDA-12.1.1 GCCcore/12.3.0
2.2.13 -CUDA-9.2.148.1 system
2.20.5 -CUDA-12.4.0 GCCcore/13.2.0
2.3.7 fosscuda/2018b
2.4.2 gcccuda/2019a
2.4.8 gcccuda/2019b
2.8.3 -CUDA-11.0.2 system
2.8.3 -CUDA-11.1.1 system
2.8.3 -CUDA-11.1.1 GCCcore/10.2.0
2.8.3 -CUDA-11.0.2 GCCcore/9.3.0
2.9.9 -CUDA-11.3.1 system

(quick links: (all) - 0 - a - b - c - d - e - f - g - h - i - j - k - l - m - n - o - p - q - r - s - t - u - v - w - x - y - z)