Designing a solution
High-performance computing (HPC) systems are available in various sizes. These solutions can start with a couple of nodes and scale out to hundreds or even thousands of nodes.
This chapter show examples of HPC systems in three different sizes to illustrate the process of their design and scaling. The examples describe solutions for small, medium, and large systems.
3.1 Interconnect and data networks
The main approach to designing an HPC or artificial intelligence (AI) system data network is to provide a high bandwidth or low ...

Get Networking Design for HPC and AI on IBM Power Systems now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.