CPRC

High Performance Computing (HPC) refers to using either supercomputers, or clusters of computers linked together, to function as a single computer. With HPC, users usually log into one central manager system and submit computing jobs using specialized software (examples include Condor, Sun Grid Engine, and Torque/Maui) which manages where and when the jobs are run.

Hotfoot Shared HPC Cluster

Hotfoot is a shared HPC cluster created under a partnership among: the departments of Astronomy & Astrophysics, Statistics, and Economics plus other groups represented in the Social Science Computing Committee (SSCC); Columbia University Information Technology (CUIT); the Office of the Executive Vice President for Research; and the School of Arts & Sciences.

The cluster runs the Torque/Maui open source scheduler software and provides 54 execute servers which provide 520 cores for running jobs. The system also includes a 51 terabyte array of working storage, shared between the departments and used by researchers for the temporary storage of their input data sets and job results.

Researchers currently using Hotfoot submit data analysis jobs using applications and languages including Matlab, R, Stata, Java, and C++. The system also supports Message Passing Interface (MPI) for parallel programming.

To request an account or to learn more about the cluster please contact Abhishek Joshi: aj2358@columbia.edu