Protocol

SHARP

Scalable Hierarchical Aggregation Protocol

Lab ready Full lab

Scalable Hierarchical Aggregation and Reduction Protocol — Mellanox/NVIDIA in-network aggregation for MPI/NCCL collectives on InfiniBand switches. Reductions inside the fabric cut bandwidth and latency for large allreduces. Requires SHARP-capable switches and enabled MPI/NCCL plugins.

Domains

ai

Layers

Tags

in-networkcollectivesibaggregation

Related protocols

InfiniBand

Lossless switched fabric with its own L2/L3 (LID/GID), subnet manager, and verbs transport (RC/UD/…). Credits, VLs, and partition keys pr…

MPI

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

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…