1 / 11

University of Virginia Alliance for Computational Science & Engineering

University of Virginia Alliance for Computational Science & Engineering. January 2008. thoughts. Goals Success metrics Non-goals Strategy Tactics Management Operations Faculty advisor BOD Resources Equipment People Support. Goals.

yoko
Download Presentation

University of Virginia Alliance for Computational Science & Engineering

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. University of Virginia Alliance for Computational Science & Engineering January 2008

  2. thoughts • Goals • Success metrics • Non-goals • Strategy • Tactics • Management • Operations • Faculty advisor BOD • Resources • Equipment • People • Support

  3. Goals • Support sponsored computational science research in Virginia • Change the culture of computation • Education/outreach/curriculum • Be the place in Virginia for HPC – be a statewide resource • Foster a multi-disciplinary ethos • Assist ITC/UVA/the schools in planning, managing, and acquiring computational resources • Rationalize high-end computing

  4. Success Metrics • Codes ported • Number of users of HPC • Science done – papers that result • Increase in research budgets associated with computational science

  5. Non-goals • Build or manage a big machine for massive macho flops

  6. Strategy • Engage university management • we need resources and policy backing • Engage the departments and schools • SOM, SEAS, CLAS, ED, all have a role, e.g., SOM bioinformatics core, ACHS, • Both in research support and in curriculum • in particular computer science • Be proactive: go to the researchers, don’t wait for them to show up. • Integrate and partner with the NSF and DOE centers

  7. Tactics • Tiger teams • Send experts to work with research groups to both parallelize their codes – and to help them plan and structure their codes • Understand our community • Gather usage statistics on clusters throughout uva • Interview users: type of problem, scope of required resources, impact if successful • Leverage national resources • Assist our users in paperwork associated with national centers, and porting their codes/scripts • Education/outreach • Bootcamps • Faculty outreach • Curriculum development and integration • Leverage local resources

  8. Management • Faculty should have major role – they are the customers • Current HPC group may be good start • Policy issues (for the queues) should be driven by faculty, not by staff • Faculty should set policy, direction, and approve initiatives • Grimshaw/Hawley lead effort

  9. Resources • Short-term • Ed and Katherine • I could provide one or two graduate students for tiger team work if they are funded • I know some very good ugrads we could try and capture • ACHS? • SOM bio-informatics core • Long term (see ITRTF document) • Full time professional staff FTE • Department faculty slots in CSE

  10. Immediate steps • Redirect Katherine and Ed to going out to the departments, interested research groups, and bootcamp participants to find tiger team projects of 3-6 month duration. Recent economics project a good model. Write up each. • Meet with them at least weekly until the projects are settled, then bi-weekly. • Ed ramps up curriculum synchronization efforts • OPPORTUNITY – CS is restructuring CS 101 • Audit/catalog clusters on grounds • Attempt to integrate under PBS? • Assess current usage patterns • Make public PC’s available for high-throughput computing via Genesis II.

  11. Computational Science in Computer Science • ----- Full --------- • Jack Davidson – optimizing compilers • Andrew Grimshaw – HPC and Grid • Anita Jones – former DDR&E, importance of computational science • Paul Reynolds – simulation & HPC • Mary Lou Soffa – optimizing and parallelizing compilers • Bill Wulf – former NAE President, importance of computational science • ---- Associate ---- • Kevin Skadron – multicore architecture • --- Assistant ------- • Sudhavna Gurumuthri – parallel IO • Kim Hazelwood – optimizing and parallelizing compilers • Marty Humphrey – Grid • Greg Humphries – parallel graphics • Wes Weimer – program annotations (including parallelism)

More Related