SRCPAC: Shared Research Computing Policy Advisory Committee

About SRCPAC

The Shared Research Computing Policy Advisory Committee (SRCPAC) is a faculty-dominated group focused on a variety of policy issues related to shared high performance computing (HPC) for primarily researchers on the Morningside, Lamont, and Manhattanville campuses. SRCPAC is also responsible for the governance of the Shared Research Computing Facility (SRCF), a portion of the data center on Morningside campus, where the jointly-acquired HPC clusters reside and are maintained by CUIT.

SRCPAC is administratively supported by the Research Computing Services group within CUIT, and the Office of Research Initiatives. The Committee additionally works closely with Columbia University Libraries and Columbia University Information Technology, particularly to offer introductory training for novice users.

SRCPAC Leadership

Current Chair

Dr. Alexander Urban-Artrith, Assistant Professor of Chemical Engineering, 2023 - Present

Past Chairs

Dr. Chris Marianetti, Associate Professor, Department of Applied Physics & Applied Mathematics, 2016 - 2023
Dr. Kathryn Johnston, Professor and Chair, Department of Astronomy, 2014 - 2016
Dr. David Madigan, Former Executive Vice President and Dean, Faculty of Arts and Sciences, 2011 - 2014

  • Dr. Jeannette Wing, Executive Vice President for Research (Chair)
  • Dr. Shih-Fu Chang, Dean, The Fu Foundation School of Engineering and Applied Science
  • Rob Cartolano, Associate Vice President for Digital Programs and Technology Services, Libraries and Information Services
  • Dr. Robert Mawhinney, Dean of Science, Faculty of Arts and Sciences
  • Gaspare LoDuca, Chief Information Officer and Vice President, CU Information Technology
  • Dr. Alexander Urban-Artrith, Chair, Shared Research Computing Policy Advisory Committee
  • Dr. Amy Hungerford, Executive Vice President and Dean, Faculty of Arts and Sciences
  • Ann Thornton, Vice Provost and University Librarian
  • Dr. Soulaymane Kachani, Senior Vice Provost, Office of the Provost, and Professor, Industrial Engineering and Operations Research
  • Dr. Cliff Stein, Interim Head, Data Science Institute
  • Victoria Hamilton, Associate Vice President of Research Initiatives and Development, Office of the Executive Vice President for Research (Staff)

SRCPAC's shared HPC Options

The SRCF pools resources from researchers across the university and negotiates centrally with vendors to amplify benefits for those purchasing hardware. It provides a professional data center and staff to install and maintain the equipment and assist researchers across the university to access these HPC clusters. To find out more information about accessing and contributing to SRCF resources, please visit: https://www.cuit.columbia.edu/shared-research-computing-facility

shopping-cart icon
Buy a server

Server purchases last five years and include ongoing support from CUIT's Research Computing Services' engineers. At times you may also purchase a portion of a server.

clock icon
Rent for a year

Rentals last a year and include support from RCS engineers.

book icon
Free Education Tier

A high-priority Education Tier on the shared HPC allows course instructors who are teaching a course or workshop addressing an aspect of computational research to request temporary access to the shared HPC resources for their students. Access will typically be arranged in conjunction with a class project or assignment.

hourglass-2 icon
Free Low-priority Tier

Researchers, including graduate students, post-docs, and sponsored undergraduates may request free access to the shared HPC on a low-priority, as-available basis. User support is limited to recorded video training sessions and online documentation only.

The University’s shared research computing clusters are not authorized to host HIPAA-protected data. The collection, storage, or transmission of Sensitive Data, as defined within the Columbia University Data Classification Policy, is strictly prohibited on Columbia's shared HPC clusters.

Attribution

We ask that published research emerging out of computations run on our shared HPC and related publications include the following acknowledgement text:

We acknowledge computing resources from Columbia University's Shared Research Computing Facility...

Recognition text for publications

Guidance for Node Purchase in Proposal

Version 4, 11.10.22

If you wish to include the purchase of nodes in the SRCPAC shared clusters in your proposal, you will want to include language in both the Facilities, Equipment and Other Resources (or equivalent) section of your proposal and in the Budget and Budget Justification. Sample approaches and links to relevant information below. You will need to adjust to the specific requests of your RFP, BAA or other. 

PLEASE NOTE THAT THE LINKS TO WEBSITES ARE FOR INFORMATION ONLY AND SHOULD NOT BE INCLUDED IN THE PROPOSAL ITSELF. 

Facilities, Equipment and Other Resources

[We/I/center name or other appropriate descriptor] will have access to the [insert name of relevant cluster], housed in the Columbia Shared Research Computing Facility, governed by the faculty-led Shared Research Computing Policy Advisory Committee (SRCPAC.)  Columbia’s Research Computing Services team administers the facility on a 24/7 basis.  [if relevant:  All members of the research team will have access to this resource.] The [insert name of cluster] has total nodes of [latest cluster specs available here: https://www.cuit.columbia.edu/shared-research-computing-facility#!#text-15697].

To perform the research outlined, [we/I/center name] intend to purchase [X] [regular, high memory, GPU, …] nodes and [Y] storage as part of Columbia’s shared high-performance computing (HPC) cluster. These clusters reside in the Columbia Shared Research Computing Facility, governed by the faculty-led Shared Research Computing Policy Advisory Committee (SRCPAC). Columbia’s Research Computing Services team administers the facility on a 24/7 basis. [if relevant:  All members of the research team will have access to this resource.]   

Every year, SRCPAC offers all researchers the chance to buy into the cluster through the purchase of nodes and storage. The purchase of a node includes peripherals (cabling, network, warranty) and support and administration for the life of the hardware (typically 5 years). The current cluster is:  [check this website for the specs if you wish to include specs https://www.cuit.columbia.edu/shared-research-computing-facility#!#text-15697.]  We anticipate pricing close to the most recent purchase round: [The most recent specific choices and costs can be found on the HPC purchase webpage .]

Budget Process

Although each sponsor may have different budgeting requirements, when budgeting for purchase of HPC nodes and active storage in a SRCPAC shared cluster on a sponsored project, it is strongly encouraged to provide as much detail in the budget justification as possible. Examples at a minimum would include: 

  • how many nodes and how much active storage are needed?
  • What shared cluster will you be using? Or will you participate in purchase of a new cluster? 
  • At what price/node? Note prices differ between regular, high-memory, GPU etc.

Vendor quotes may be required as backup documentation, should the sponsor request it. For purposes of the proposal, we recommend using the latest pricing which you can find.  You may wish to state explicitly that if selected for funding, a revised quote can be provided. The most recent specific choices and costs can be found on the website under [The most recent specific choices and costs can be found on the HPC purchase webpage.]

Note that the price history should be used to give you an estimate using the most recent information available on costs. If your proposal is awarded, the actual costs may differ. The sponsor may request an updated quote.

PLEASE NOTE THAT THIS IS ACTIVE STORAGE THAT IS USED WHILE COMPUTING.  IT DOES NOT FULFILL THE PURPOSE OF PERMANENT OR ARCHIVAL STORAGE.  For storage options are outlined on the data repository webpage.

Several computer nodes will be purchased to exclusively support the code development and computation associated with the research in this grant, and these nodes will be managed by Columbia within our Shared Research Computing Facility. Each node will approximately consist of a dual socket mother board containing two 16 core Intel Xeon processors. Two standard nodes, each having 192GB of RAM, will be purchased during the first year, with an estimated cost of approximately $7,850/node (see attached quote from previous purchase round), accounting for the $15,426 budgeted during year 1. A high memory node, containing 768 GB of RAM, will also be purchased, accounting for the $11,658 budgeted during year 2. Finally, approximately 11 TB of storage space will be purchased at an estimated cost of approximately $350/TB, accounting for the $4,341 budgeted during year 3.

Join Us!

users icon
Stay in the loop!

SRCPAC meets every Fall and Spring semester for approximately 60 to 90 minutes, with select faculty, administrators, and leadership presenting updates pertaining to the University's shared research computing infrastructure. All Columbia faculty, research scientists, postdocs, students, and administrative staff are welcome to attend meetings. Sign up for the SRCPAC mailing list to be informed about meetings and announcements on research computing developments.

Archive

SRCPAC Charter

Excerpt from the SRCPAC Charter, November 9, 2011:

"The Shared Research Computing Policy Advisory Committee (SRCPAC) will be a faculty-dominated group focused on a variety of policy issues related to shared research computing on the Morningside campus. As the use of computational tools spreads to more disciplines to create, collaborate, and disseminate knowledge, there is a commensurate rise in the costs of establishing and maintaining these resources. Shared resources have proven to leverage those available to individuals or small groups, but require careful consideration of the policies governing the shared resource and the basis of the operating model.

While final authority and responsibility for such policies customarily rests with the senior administrators of the University, it is vital that the research faculty examine and recommend the policies and practices they deem best suited to accomplishing the research objectives."

SRCPAC Proposed Mandate

Materials from RCEC and SRCPAC Meetings

Completed SRCPAC Subcommittees & Working Groups

  • ​​​​​​Intercampus Subcommittee
  • Columbia Survey Working Group
  • Cloud Subcommittee
  • External Peer Survey Working Group
  • Hotfoot HPC Operations Committee
  • Manhattanville Liaison Working Group
  • Research Storage Working Group
  • Training Subcommittee
  • HPC Operating Committee

Publication Outcomes

Research conducted on our high performance computing machines has led to many peer-reviewed publications in top-tier research journals. To view citations for these publications please visit:

To report new publications utilizing one or more of these machines, please email [email protected].