Protocol

RCCL

ROCm Collective Communications Library

Lab ready Full lab

AMD ROCm Collective Communications Library — the ROCm counterpart to NCCL for GPU collectives. It uses Infinity Fabric, PCIe, and RDMA paths with an API largely mirrored from NCCL for portability. Topology and RCCL env knobs are the usual performance levers.

Domains

ai

Layers

application

Tags

collectivesamdrocmallreduce

Sources

AMD RCCL documentation

Related protocols

NCCL

NVIDIA Collective Communications Library implementing allreduce, broadcast, and related GPU collectives over NVLink, PCIe, and RDMA. Topo…

RDMA Verbs

Kernel-bypass networking model: queue pairs, memory regions, and work requests move data between registered buffers with minimal CPU. Com…

MPI

Message Passing Interface — portable parallel messaging with point-to-point and collective operations across ranks. Implementations map t…