Overview

At the core of CORE lies the multi-core!

... and that multi-core is housed and administered by the Center for Simulation and Modeling (SAM), and supported through this web-site (CORE). To read more about SAM and CORE see About.

The central High Performance Computing cluster at the University of Pittsburgh is called Frank. Additionally, we have MPI and HTC clusters which serve more specific needs. If you are interested in getting an allocation, please contact us for further information, or (faculty only) submit a short proposal for allocation.

Frank, MPI, and HTC are all maintained and administered by the Center for Simulation at Modeling (SAM) and CSSD.

FRANK

Frank is an heterogeneous system featuring unique high-end computational capabilities, including high capacity multi-core shared memory systems, and NVIDIA GPUs.

Frank currently consists of a total of 7808 CPU cores distributed over 384 nodes. See the system architecture page for more details.

MPI

MPI is a homogeneous system designed to run massively parallel multi-node calculations. MPI currently contains 22 10-core Intel Haswell processors E5-2660 v3 @ 2.60GHz (2 sockets per node).

HTC

HTC is a homogeneous system designed for high throughput computing. HTC currently contains 16 8-core Intel Haswell processors E5-2630 v3 @ 2.40 GHz (2 sockets per node).