Reviews of MSI and Recommendations for its Future

Transition Planning Committee

On February 19, 2008, Tim Mulcahy, University of Minnesota Vice President for Research set up a committee to plan MSI's transition from the Institute of Technology to the Office of the Vice President for Research. The members of the committee are:

The charge of the committee is:

In April 2007, internal and external advisory committees commissioned with advising the University administration on the future of the Minnesota Supercomputing Institute (MSI) recommended substantial changes to the current organization, all intended to facilitate the University's move into the top ranks of research institutions. The committee reports are available [below].

The committee recommendations have been accepted in principle by the administration. As recommended by the External Advisory committee, it has been determined that administrative responsibility for the MSI will transfer from the Institute of Technology to the Office of the Vice President for Research, where it will form the core of a larger campus-wide scientific computing support organization. The transfer is scheduled to be effective July 1, 2008, corresponding to the start of the new fiscal year, and Andrew Odlyzko's stepping down from his current role as MSI Interim Director and Director of the Digital Technology Center. This organizational move is intended to facilitate implementation of changes needed to preserve and enhance MSI's role as a vital computational resource across the entire University, without diminishing its already strong research contributions. Effective with this transfer, MSI will be renamed the Minnesota Research Computing Center, as suggested by the External Advisory Committee, to reflect its centrality to the research mission and ambitions of the University.

In order to accommodate this transition with a minimum of disruption to the research groups that rely on the resources and services of the MSI, I am appointing this Transition Planning Committee (the "committee") and charging it with the development of a plan for the transition. It is expected that it may take several years of evolution to reach a new steady-state incorporating the recommendations of the advisory committees and alignment with ongoing developments in national funding opportunities and with the University's strategic priorities. Nevertheless, the University needs to start planning now to address the short and medium-range issue challenges to be confronted during this upcoming transition. Therefore I would ask the Transition Planning Committee to recommend a transition plan addressing each of the following issues:

  1. How should the MSI be structured to most effectively meet the needs of the University research community and to support research in high-performance computing?
  2. What services should MSI offer, how should they be made available and how should they be configured across campus?
  3. What relationship should MSI have with other research computing support units?
  4. How should future infrastructure needs be identified and prioritized?
  5. What are the most promising initiatives (internal and external) for MSI to pursue?
  6. What should the roles and responsibilities of the Director be and what are the relative advantages and disadvantages of selecting a new Director from campus versus following a national search?
  7. Based on the answer to question 6, develop an outline for a job description for the Director's position.

Although the timeline is aggressive, I would like the committee to submit a written report summarizing their recommendations by April 15, 2008, in order to allow broad discussion of the recommendations in time to take any necessary action before MSI moves to OVPR, and to guide decisions with regard to the appointment of the next director.

You can read the full memo of the Charge to the MSI Transition Planning Committee.

Reviews of MSI and Recommendations for its Future

During 2006 and 2007, the Supercomputing Institute underwent an evaluation of its structure and direction with the goal of determining the best course for its future. The Institute established two internal long-range planning committees as vehicles for this process: one focusing on computation in the biosciences and the other focusing on computation in engineering, math, and the physical sciences.

To assess the needs and desires of Institute principal investigators, a survey was taken in December 2006, and 91 faculty responded. The survey asked the principal investigators to evaluate the importance to their research of the Supercomputing Institute's core resources, software, and laboratories. The results showed strong support for the Institute and extensive use of high-level computing in research work. Researchers were also asked to submit suggestions concerning the future structure of the Institute. A report summarizing the results of the survey and of the extensive investigations and deliberations of the planning committees was sent to University of Minnesota administration on March 7, 2007 and was distributed to Institute principal investigators at the same time.

Complementing the efforts of the internal committees, on April 2-3, 2007, the Institute was visited by an External Advisory Board composed of three eminent experts in computational research:

Prof. Daniel A. Reed, University of North Carolina, Chair
Prof. John Quackenbush, Harvard and Dana-Farber Cancer Institute
Dr. Jeremy Kepner, MIT Lincoln Laboratory

The Board met with Institute committee members, Institute staff, principal investigators, and University administration members; and they reviewed the structure and mission of the Institute. Board members were overwhelmingly positive about the need for the Institute at the University of Minnesota and stressed that a strong, well-funded Institute would support the University's goal of becoming a top research university. They also recommended making the Institute the core of a considerably larger and more comprehensive scientific computing support organization.

External Advisory Board report
External Advisory Board detailed presentation deck
Internal long-range planning committees' report