High Performance Computing
The ability to model multiphysics problems at multiple temporal and spatial scales is increasingly needed to understand the performance of next generation reactors, fuel cycles, and materials. High performance computing plays a central role in these efforts. The need to perform massively parallel computations in reactor physics and thermal hydraulics is now the norm, not the exception. The properties of next generation reactor materials are often tested and refined through simulation as well.
The center for High Performance Computing at The Colorado School of Mines maintains a 10,496 core IBM BlueGenQ supercomputer as well as a scalable Penguin cluster with 1,920 Intel cores and 2,304 Nvidia GPU cores. Both systems are available to faculty across the University, and are actively used in research by faculty in the Nuclear Science and Engineering program.
The Nuclear Science and Engineering program maintains a student computing laboratory for instruction in the MCNP family of codes as well as RELAP, SCALE, SERPENT, COMSOL, Solidworks and Matlab. The laboratory is equipped with 10 Mac Pro workstations (five eight-core machines and five 12-core machines) that are linked to form a cluster with which students can learn MPI and MPICH. The computer lab also contains a Windows-based tomographic reconstruction workstation.