Glossary¶
- PSC: the Pittsburgh Supercomputing Center, home of the Neocortex system and team.
- Allocation: a group of resources from the PSC systems, Neocortex and Bridges-2.
- Service Unit (SU): is a term used for tracking the allocation and use of resources in the system in terms of the resources being used in hours. For example, when using the CS-2 machines, the amount of total CS SU starts decreasing from the time you start a Slurm job including the CS machine and the prompt appears; and it ends when the job finishes.
- CS: abbreviation of the Cerebras machine names. For example, CS-2.
- SDF: abbreviation of the Hewlett Packard Enterprise Superdome Flex server 280, as in SuperDome Flex.
- Model Zoo (modelzoo): this is the name Cerebras gives to their repository of reference model examples.
- Slurm is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small Linux clusters. It is the main job scheduler used on both the Neocortex and the Bridges-2 systems.