The simulations performed by CCRG researchers require the use of the state-of-the-art High-Performance Computing (HPC) systems. The Center hosts and manages several single-owner project-specific computing facilities:

• "gravitySimulator," a special-purpose 32-node supercomputer that achieves speeds of 4 Tflops in solving the gravitational N-body problem. gravitySimulator was purchased in 2004 through an NSF MRI award (AST-0821141). More information about this cluster can be found here.
• Three stand-alone GRAPE-6 computers.
• "newHorizons," a 85-node Linux cluster with dual-processor (dual-core) per node, Infiniband interconnects and a total of 1.4 Tbytes of on-board memory. The design of newHorizons was optimized for numerical relativity simulations to achieve speeds up to one Tflop while utilizing large amount of memory and being able to work with large data sets. NewHorizons was purchased in fall 2007 through an NSF MRI award (PHY-0722703). See related PR and article.
• A 16-node Linux cluster with dual processor (quad-core) per node, Infiniband interconnects and a total of 0.7 Tbytes of on-boarch memory. This is a new experimental cluster for numerical relativity simulations, which will be connected to "newHorizons" bringing up the system to 1.2 Tflops of overall peak performance. It is funded through a NSF SCREMS award (DMS-0820923) and may become available in spring/summer 2009.
• Additional special hardware facilities, including: 22 Tbytes of common storage unit, a large monitor for scientific visualization and several computer servers.

In addition, we also make use of larger Terascale and Petascale computing resources available at TeraGrid and Open Science Grid sites. Some of our largest simulations are performed at TeraGrid sites, such as TACC, NSCA, etc.