SLURM
Definition
SLURM (Simple Linux Utility for Resource Management) is an open-source, highly scalable cluster management and job scheduling system used by many of the world’s supercomputers and Linux clusters. SLURM allocates compute resources (nodes, CPUs, GPUs, memory) to jobs, provides a framework for starting and monitoring work, and manages a queue of pending jobs. It supports fair-share scheduling, preemption, job arrays, and heterogeneous job steps.
