High performance computing @Oberlin College

SCIURus is the high-performance computing cluster for Oberlin College, purchased using funds from the National Science Foundation. Installed in May 2015 and housed in the Oberlin Science Center, the cluster consists of 36 Dual Intel Ten Core Xeon E5-2650v3 (Haswell) processors, each with 2.3GHz and 256 GB RAM, for a total of 720 compute cores. It also contains a single GPU node (NVIDIA Tesla K40 PCI-E 12GB GDDR5 Passive Single GPU) and a single Intel Xeon Phi node (Xeon Phi 5110P coprocessors). All nodes are connected by 10 Gb ethernet. The total storage of the cluster is 192 TB. The cluster is managed using CentOS v6.

SCIURus is a resource to advance the research and teaching goals of Oberlin College, and its usage and policies are managed by the Oberlin College HPC Committee, which currently consists of the following Oberlin College faculty and staff:

Mike Moore, Associate Professor of Biology, PI and Chair of the Committee (mmoore@oberlin.edu)
Chris Mohler, System Administrator, OCTET (cmohler@oberlin.edu)
Manish Mehta, Professor of Chemistry and Biochemistry, Co-PI
Matt Elrod, Professor of Chemistry and Biochemistry, Co-PI
Aaron Goldman, Assistant Professor of Biology, Co-PI
Rob Owen, Assistant Professor of Physics and Astronomy, Co-PI

Potential users should fill out the online application form. You will be required to log in using your ObieID, if not already logged in.

Where does the name of the cluster come from?
SCIURus has a triple meaning. The first five letters of SCIURus are an acronym for “SuperComputing in Undergraduate Research”, and the first three letters correspond to the first three letters of the word science. The name of was also chosen because it is the official Latin-based taxonomic name for squirrels (the genus Sciurus), in reference to the white squirrels that are often found in Oberlin and which are the unofficial mascot of Oberlin College.

Usage Policies

Usage of SCIURus is subject to the following policies:

(1) Users must be Oberlin College faculty, affiliated research staff, or must be officially affiliated with Oberlin College faculty. Hence, any Oberlin College student wishing to be a SCIURus user must be working with an Oberlin College faculty sponsor as part of an official research or class project. Off-campus users must be official collaborators with Oberlin College faculty (meaning that any work performed by outside collaborators should normally lead to publications with Oberlin College faculty co-authors), and must be sponsored by an Oberlin faculty member (i.e., the Oberlin faculty collaborator must fill this form out on behalf of the collaborator).

(2) SCIURus may not be employed for personal use, nor may it be employed for commercial or any other for-profit use.

(3) All users must comply with software licensing agreements, and Oberlin faculty should take care to ensure that any users under their sponsorship also comply with licensing agreements.

(4) SCIURus is a shared resource. Users should refrain from routinely submitting extremely large jobs to the cluster during periods of peak demand, and should not use the storage array for long-term storage of large files. Long-term data storage is the responsibility of each user.

(5) Users will be required to fill out annual reports detailing their SCIURus usage and any publications resulting from usage. All publications resulting from use of SCIURus should acknowledge the cluster and the number of the NSF grant that funded the cluster: MRI 1427949. Suggested language is as follows: “(Analyses) were conducted on SCIURus, the Oberlin College HPC cluster (NSF MRI 1427949)”.

(6) The HPC Committee reserves the right to restrict usage to any user who is deemed to be in violation of any of the above policies.

The usage and policies of SCIURus are managed by the Oberlin College HPC Committee, which currently consists of the following Oberlin College faculty and staff:

Mike Moore, Associate Professor of Biology, PI and Chair of the Committee
Manish Mehta, Professor of Chemistry & Biochemistry, Co-PI
Matt Elrod, Professor of Chemistry & Biochemistry, Co-PI
Aaron Goldman, Assistant Professor of Biology, Co-PI
Rob Owen, Assistant Professor of Physics & Astronomy, Co-PI

Interested in using SCIURus?

SCIURus is a tool for research and education at Oberlin College, and as such its use is limited to Oberlin faculty, staff, students, and officially affiliated collaborators who are conducting research or course-related activities that require HPC resources.

More details on usage policies can be found on the user application form, which all potential users must fill out.

Additional information about the cluster can be found here.

Print Friendly, PDF & Email