#
Types of compute job
Users can run a number of different types of job via the research environment scheduler, including:
- Batch jobs; single-threaded applications that run only on one compute core
- Array jobs; two or more similar batch jobs which are submitted together for convenience
- SMP or multi-threaded jobs; multi-threaded applications that run on two or more compute cores on the same compute node
- Parallel jobs; multi-threaded applications making use of an MPI library to run on multiple cores spread over one or more compute nodes
The research environment job-scheduler is responsible for finding compute nodes in your research environment to run all these different types of jobs on. It keeps track of the available resources and allocates jobs to individual groups of nodes, making sure not to over-commit CPU and memory. The example below shows how a job-scheduler might allocate jobs of different types to a group of 8-CPU-core compute nodes:
