A brief overview of the Halcyon Cluster’s computer architecture should prove helpful in gaining an understanding of how the components are managed. Its high-performance computing infrastructure and extensive software are controlled and monitored by a Maui scheduler and a Torque resource manager. Together, these tools manage the execution of jobs by analyzing their resource requests (i.e., time, memory, speed, and software) in relation to other operations within the greater environment. Accordingly, jobs are optimally queued to run on the cluster.
The cluster is comprised of 40 nodes. Each node has a 2 x 2.8GHz hex-core Intel CPU for a total of 12 cores with a capacity of 96GB memory per node. Not only does its supercomputing technology facilitate cutting-edge research, the cluster also serves to prepare students for the field of bioinformatics. Any approved computer within the University system can be used to access the cluster.
All jobs are written in bash shell, a Unix platform that enables command-line interfacing and batch scripting; additionally, a variety of programming languages (e.g., Perl, Python, and Ruby) can be incorporated.