The Condor system is designed to maximize throughput rather than performance; hence, the system is often referred to as a High Throughput Computing (HTC) system. The system is also opportunistic in that it is available on a first-come-first served basis only when the systems are not being used by their primary owners. The condor cluster at FSU is constructed out of mostly donated and fully depreciated hardware. The system is specifically tailored to support large batches of serial jobs where overall throughput is preferred over performance. For example, this cluster is well suited to run numerous serial jobs simultaneously. A special purpose log in node is available for condor submissions. The Lustre storage system is mounted on the condor login node.
Condor is our High Throughput Computing service. HTC allows users to leverage many computing nodes to accomplish a single parallelized task. Popular use-cases include Matlab and other applications.
The Condor Cluster is available to all FSU faculty and students on a first-come, first-served (or first-job-submitted, first-job-processed) basis.
|Memory per Core||2gb to 8gb|
|Storage System||Lustre (Terascala)|