Peter Marshall, ITS, UWO <peter@julian.uwo.ca>
This summary is based primarily on a meeting with Bruce Attfield, 1997-08-01.
- Proposal, primarily to the Canada Foundation for Innovation (for startup funds) and to the Canadian Government (for continuing funds), to establish and maintain a distributed high-performance computing backbone for Canada.
- Network, computation sites (including support personnel), application development support and education programs to be funded.
- Initial proposal for $100 million over 5 years, but it may turn out to be $140 million over 7 years -- this is intended to be a long-term project that anticipates strong government and industry funding for its entire life -- no selling of "cycles" to help fund (that model has never worked).
- Phase I: get funding Phase II: upgrades to existing centres (like UWO?) Phase III: still not well defined -- setting direction now, probably based on the experience in Phase II -- does the distributed model work?
- Still lots of definition work to be done.
- Primarily participation and high-level backing from UWO to help ensure that this proposal is the one that moves forward and gets the funding.
- Secondarily that we contribute $10,000 to the report-writing and lobbying effort that will be required to get the Collaboratory established over the next year or two. Total budget for this phase about $300,000
- $50,000 seed money from NRC, and a total of $210,000 has been collected so far from all sources indicates strong monetary support from the Canadian University and research community.
- The Collaboratory may be a source of funds to replace a (then) aging Cray J90se in two or three years -- we no longer have a replacement budget for a future high-performance computing machine. This may provide a large computational resource locally with some sharing required but local access and local expertise combined with a bigger machine than we could afford on our own may win out.
- Possible funding for the Proposed Computational Science masters program. One of the main deliverables of the Collaboratory is an educated and experienced group of researchers. I'm told that industry is crying out for such specialists. (Suggestion from Bruce Attfield -- that we consider making this program coop to get industry involved early.)
- We need to be represented as an active stakeholder in order that these benefits move beyond the proposal scale and into the support of the facilities and programs at UWO.
- Need to hold a high-level meeting by the end of August probably with Bruce Attfield <attfield@fox.nstn.ca> and get someone from Western actively working on making this proposal happen. This might be a key researcher or it might be someone from the administration. Queens' VP Research (Suzanne Fortier) is very active in the committees.
- Perhaps, because of our proposed Computational Science program, Western might take the lead by taking responsibility for the Education and Training Committee.
- Assuming a favourable outcome of a meeting, that UWO become a stakeholder in this project by contributing $10,000 to moving the proposal forward.
- We look into including a coop component into the Computational Collaboratory program.