Introduction
Ever wondered what happens when a mid-sized company decides to invest significant cash in a new high-performance computing setup, to better handle present and future application performance?
How does it decide on hardware infrastructure, for example, and, just as importantly, how is software optimised to leverage the strength of multi-core CPUs, where eight processing cores can be implemented in a small form-factor IBM BladeCenter?
HEXUS was invited to IBM's High-Performance Computing Joint Competence Center (HPCJCC) in Montpellier, France, to grab a first-hand look at large-scale, cluster-based solutions at work, how they're benchmarked, and the reason(s) why mid-sized companies choose to come here to evaluate the high-performance-computing hardware required for their particular apps.
