High-performance computing clusters rely on an optimized Message Passing Interface (MPI) and predictable low-latency networking to support high performance parallel software workloads. As organizations continue to build out their HPC configuration, MPI and networking are integral to achieving a high-performance, low-latency, production-quality, and cost-effectiveness solution.
In this webinar, Platform Computing, Arista Networks and Solarflare will provide results from MPI and 10GbE latency tests and 10GbE cost analyses, and a reference architecture for building a cost-effective, low-latency 10GbE HPC solution.
|