CIERA has exclusive access to two high-performance computer clusters. These two clusters are used not only for running simulations and doing analysis for research, but serve as a training ground for undergraduate students in HPC computing and computational research. In total, the two clusters provide access to 1,344 compute cores. Additional HPC resources, including high-memory nodes, are available through Northwestern’s Quest cluster.
The NSF-funded Grail cluster is available for use by CIERA researchers doing work related to gravitational waves. This computer cluster provides 37 compute nodes for general-purpose high-performance computation with, as well as 2 special-purpose nodes with Graphics Processing Units (GPUs) appropriate to scientific computing which allow accelerated parallel computations. The CIERA cluster is available for use by any CIERA researcher for astronomy-related research, and has a total of 9 standard nodes.
Almost all nodes have dual 14-core CPUs (Intel(R) Xeon(R) Gold 6132 (Skylake) CPU @ 2.60GHz), for a total of 28 cores per node. The standard compute nodes and GPU nodes have 192 GB of RAM per node. The nodes are interconnected with Infiniband FDR. Grail’s two additional GPU nodes contain dual NVIDIA Tesla V100 cards; each card has two NVIDIA V100-PCIE-16GB GPUs with 16GB of RAM per GPU. The clusters also contain significant storage space for computations and simulation results: a total of 85 TB of high-speed disk storage directly connected to the cluster in a RAID configuration.
Grail and CIERA Clusters
The Northwestern’s Research Computing group houses and maintains the Grail and CIERA clusters alongside its Quest High-Performance computer. Their group has excellent space, cooling, and power (with UPS support) to support the system.
Funding for the Grail computer cluster was provided by the National Science Foundation’s Major Research Instrumentation (MRI) program, grant PHY-1126812.
CIERA researchers looking for details on how to access and use the Grail and CIERA clusters should refer to the wiki page.