high performance cluster (HPC)

High PerformanceCluster(HPC) is a flexible alternative to mainframes and supercomputers. HPC systems set up as clusters are scalable, easy to manage and have high availability.

An HPC cluster distributes its applications to the available processors using middleware. The HPC cluster is managed by a front-end processor( FEP), which controls the individual computer nodes via a switch. The partial results of the compute nodes are delivered to the front end. HPC clusters can consist of a few processors, but they can also comprise several thousand. The individual computing nodes of an HPC cluster may be equipped with their own memories or connected to high-speed networks via appropriate network cards.

HPC systems are high- performance computer networks with short latency times, with which complex calculations and simulations can be executed in fractions of a second.