High Performance Computing

What's the problem?

You have computational needs that exceed the capabilities or availability of your desktop PC. You have either a single large problem to solve, or dozens, hundreds, or thousands of smaller related problems to solve.

Why should I care?

You don't need to tie up your desktop PC with long-running computational work when you could be writing papers, developing programs, handling email, etc. You may not have the funds to purchase a dedicated desktop system for simulations, or would prefer to use available funds for other equipment or for personnel time. Running jobs in PC labs runs the risk of others rebooting the PC during the middle of a run, and reduces the facilities available to the students for their own work.

What's your solution?

Use the CAE High-Performance Computing Cluster. With a total of 98 CPUs available 24/7 and another 70 CPUs available during late nights and on weekends, cluster systems contain up to 8 GB memory, 500 GB temporary disk space, and 4 CPUs per system. Some jobs can be farmed out to dozens of smaller PCs in an overnight cluster, perfect for running large numbers of small jobs. Large-scale jobs that may benefit from parallel processing can combine up the capabilities of up to a dozen systems into a larger parallel computer with corresponding performance boosts. The cluster's queuing system allocates resources to each job to ensure that each user's job is run independently and efficiently on dedicated processors.

Where can I find more information?

Contact Mike Renfro (renfro@tntech.edu, 372-3601, Clement Hall 314), or see this documentation.