High Performance Computing (HPC)
High performance computing clusters are designed to support a broad range of workloads that cannot be supported by traditional workstations. ARCS currently offers one HPC cluster — the Novus cluster — providing shared computing resources open to the entire OSU research community, including classroom instruction with needs in data science computing.
Cluster Access
Each cluster provides three levels of tiered access.
Standard
Free for all students and researchers at OSU that provides access to 10 shared general compute nodes with 100GB of file storage on the HPC cluster.
Shared
In addition to the standard tier, you may access additional compute nodes that are owned by research groups. Note: jobs submitted, in the shared queue might get killed if the job was running on a research group compute node when they submit a new job.
Condo
Condo investment into the HPC cluster. A researcher or group can opt to purchase their own compute node(s) to be added into the HPC cluster. The ARCS team will work with the researcher to purchase compute nodes that best meet their needs. ARCS will then work with our vendor to purchase, install, administer, and maintain the node(s) for free for five years. 100 percent of these compute node(s) will be made available to the researcher on-demand through separate partitions in the scheduling system. When the compute node(s) are not in use, their compute resources will then be made available to the general shared queue on the cluster. It is expected that the researcher will refresh their compute node(s) at the end of the five-year term.
Cluster Storage
Each account on the cluster will get a default 100GB of file storage. Additional storage can be purchased if needed. In addition, there is scratch file space open to everyone for larger data sets and data outputs files can be stored. The scratch space meant for temporary storage only, any files that have not been accessed in 90 days will be automatically deleted.
Acknowledgment
Please acknowledge your use of the Advanced Research Computing Services High Performance Computing environment.
Maintaining an HPC environment is expensive, and the Advanced Research Computing Services team asks that you acknowledge your use of the ARCS cluster in papers or other publications of research which benefited from the use of this resource. These acknowledgments can assist us in showing the value of this resource to the campus and help in getting funds for the continued maintenance and/or future expansion.
To acknowledge your use of the cluster, we request that you use the following wording:
"The authors acknowledge the Advanced Research Computing Services (ARCS @OSU) supercomputing resources (https://arcs.oregonstate.edu) made available for conducting the research reported in this paper."
It would also be helpful if you let us know whenever you have acknowledged the cluster in any publications.