Alternatives and similar repositories for nccl

Users that are interested in nccl are comparing it to the libraries listed below

• NVIDIA / ib-traffic-monitor
  A TUI-based utility for real-time monitoring of InfiniBand traffic and performance metrics on the local node
  ☆63Updated last month
• facebookresearch / param
  PArametrized Recommendation and Ai Model benchmark is a repository for development of numerous uBenchmarks as well as end to end nets for…
  ☆156Updated last week
• ROCm / rccl-tests
  [DEPRECATED] Moved to ROCm/rocm-systems repo
  ☆86Updated 2 weeks ago
• Infrawaves / DeepEP_ibrc_dual-ports_multiQP
  Aims to implement dual-port and multi-qp solutions in deepEP ibrc transport
  ☆73Updated 9 months ago
• Azure / msccl-executor-nccl
  ☆47Updated last year
• uxlfoundation / oneCCL
  oneAPI Collective Communications Library (oneCCL)
  ☆254Updated last week
• aws / aws-ofi-nccl
  This is a plugin which lets EC2 developers use libfabric as network provider while running NCCL applications.
  ☆204Updated this week
• deepspeedai / DeepSpeed-Kernels
  ☆71Updated 10 months ago
• google / nccl-fastsocket
  NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.
  ☆122Updated 2 years ago
• Azure / msccl
  Microsoft Collective Communication Library
  ☆66Updated last year
• intel / torch-ccl
  oneCCL Bindings for Pytorch* (deprecated)
  ☆104Updated last month
• microsoft / NPKit
  NCCL Profiling Kit
  ☆152Updated last year
• NVIDIA / MagnumIO
  Magnum IO community repo
  ☆109Updated 2 months ago
• abcdabcd987 / libfabric-efa-demo
  ☆77Updated last year
• microsoft / ark
  A GPU-driven system framework for scalable AI applications
  ☆124Updated last year
• NVIDIA / mlperf-common
  NVIDIA's launch, startup, and logging scripts used by our MLPerf Training and HPC submissions
  ☆35Updated 5 months ago
• openucx / torch-ucc
  pytorch ucc plugin
  ☆23Updated 4 years ago
• NVIDIA / cloudai
  CloudAI Benchmark Framework
  ☆83Updated this week
• gpudirect / libgdsync
  GPUDirect Async support for IB Verbs
  ☆135Updated 3 years ago
• ROCm / rocm_bandwidth_test
  Bandwidth test for ROCm
  ☆75Updated last week
• meta-pytorch / torchcomms
  torchcomms: a modern PyTorch communications API
  ☆330Updated this week
• meta-pytorch / triton-cpu
  An experimental CPU backend for Triton (https//github.com/openai/triton)
  ☆49Updated 5 months ago
• ROCm / rccl
  [DEPRECATED] Moved to ROCm/rocm-systems repo
  ☆410Updated this week
• ROCm / rocSHMEM
  [DEPRECATED] Moved to ROCm/rocm-systems repo
  ☆144Updated this week
• NVIDIA / nvshmem
  NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…
  ☆462Updated last month
• microsoft / tokenweave
  Accepted to MLSys 2026
  ☆70Updated 2 weeks ago
• vllm-project / flash-attention
  Fast and memory-efficient exact attention
  ☆114Updated this week
• openucx / ucc
  Unified Collective Communication Library
  ☆291Updated last week
• amd / ZenDNN
  ☆137Updated last week
• Oneflow-Inc / dfccl
  ☆26Updated 11 months ago