What is an HPC?
An HPC, or High Performance Computer, is a logical computer made up of many physical compute nodes and sometimes a storage node.
What is the benefit of an HPC?
The obvious benefit is…more resources, more power! But an equally important benefit is software optimized to run jobs in parallel. And another benefit is resources management, provided through a workload manager.
What is Parallel Computing?
It is the ability to split up a large processes into smaller tasks, that are solved simultaneously across compute nodes.
What is a Workload Manager?
When a resource is highly coveted, it can become difficult for others to get usage time. A workload manager allows for anyone to submit a job, and it be placed in a queue that is first come first serve.
What is Scratch Space?
Scratch Space refers to the temporary working directory that a job runs on. It is the largest disk space, stored on the storage node. Usually, scratch space is accessed by /scratch or $scratch, but in our cluster, the /home/username directory is your scratch space. Just keep it clean!