This page is a primer on the HPC. Below, you will find information about what the HPC is, how to connect, and what kinds of work you can do with the system.
What is the HPC?
The High Performance Computing Cluster at FSU is a collection of servers designed to process long-running, compute-intensive batch jobs. The cluster consists of a large number of compute nodes connected by a high performance Infinband network. This type of system is commonly referred to as a Beowulf cluster.
What is it used for?
The HPC is used for long-running batch jobs that require intensive compute resources. Jobs can be monitored, but are not interactive. Compute jobs running in the HPC can operate in parallel using popular frameworks like OpenMP and MPI.
Many users write and/or compile their own software to run on the HPC, for which we provide a number of tools and libraries to support. Other users can run jobs in general purpose applications, already compiled and installed on the HPC, such as MATLAB.
How to Connect
To use the HPC, connect via SSH to our login nodes:
MOAB is a job manager, and it is the tool that users interact with to submit and manage jobs in the HPC queues.
Torque is a job scheduler, and is the back-end application that provisions resources. Most users will never interact with Torque directly.
Jobs and Queues
Anytime a user wishes to run a job on the HPC, he or she will create a MOAB job, and submit that job to one of the processing queues. Jobs in the queue run in the order they are submitted. This means that your job may not execute right away, but it allows us to provide HPC resources to users in such a way so that all jobs eventually run.
We provide many queues for different groups at FSU. Any faculty or student with an RCC Account can use submit jobs to our general access and backfill queues. For those research groups who have purchased resources on the HPC, we provide dedicated queues with enhanced access to resources in the HPC cluster.