A High-Performance Superpipeline Protocol for InfiniBand - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2011

A High-Performance Superpipeline Protocol for InfiniBand


InfiniBand high performance networks require that the buffers used for sending or receiving data are registered. Since memory registration is an expensive operation, some communication libraries use caching (rcache) to amortize its cost, and copy data into pre-registered buffers for small messages. In this paper, we present a software protocol for InfiniBand that always uses a memory copy, and amortizes the cost of this copy with a superpipeline to overlap the memory copy and the RDMA. We propose a performance model of our protocol to study its behavior and optimize its parameters. We have implemented our protocol in the NewMadeleine communication library. The results of MPI benchmarks show a significant improvement in cache-unfriendly applications that do not reuse the same memory blocks all over the time, without degradation for cache-friendly applications.
Fichier principal
Vignette du fichier
main.pdf (286.51 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

inria-00586015 , version 1 (14-04-2011)


  • HAL Id : inria-00586015 , version 1


Alexandre Denis. A High-Performance Superpipeline Protocol for InfiniBand. Euro-Par 2011, Aug 2011, Bordeaux, France. pp.276-287. ⟨inria-00586015⟩
331 View
1320 Download


Gmail Facebook X LinkedIn More