Glossary

cluster

Computer cluster is in the broad sense the terminology for a supercomputer, consisting of a set of connected computers working together so that they can be viewed as a single system.

node

Nodes are components of a cluster and analogous to the computers we use every day. A supercomputer consists of a number of nodes that perform computations and run an own instance of the operating system.

partition

Nodes are logically grouped into partitions. See resources to examine what partitions currently has available.

processor

component of each node. Each node has a number of processors that are analogous to the CPU (Central processing unit) in a personal computer.

core

component of each processor. The actual computing entity.

corehours

Corehours refers to the number of processor units (cores) used to run a simulation multiplied by the duration of the job (t) in hours. Corehours (c) can be calculated if the number of nodes (n), the duration in hours (t) and the number of cores per node (cpn) are known: c = n*cpn*t

SLURM

SLURM or Simple Linux Utility for Resource Management is an open-source cluster management and job scheduling system.