Date

Attendees

Goals

Discussion items

TimeItemWhoNotes
Successful XSEDE Proposal Webinar
    • Keep it short
    • If you’re using code you wrote, explain why
    • Provide detail about data to be stored, and amount and expected growth of data to be stored
      • III. Application Efficiencies:
        • Why you chose the specific resources?
        • If you have your own code, how does it scale – what’s the performance?
        • Efficiency of storage algorithms and protocols
        • Describe and estimate expected growth
      • IV. Computational Research Plan
        • Explicitly describe the problem cases you will examine
        • Explicitly describe the typical use-cases that the gateway supports and the type of runs that you expect users to make
      • V. Justification of USs, TBs
        • Straightforward math problem if you were clear enough in sections II-IV
        • For each research problem, calculate the SUs required based on runs defined in IV and timings in III.
        • Community gateway-type proposals:
          • First year is trickier to calculate SUs / requires bootstrapping
          • Beyond 2nd year, estimate total usage based on 1st year patterns
      • VI. Additional Review Considerations
        • Ability to complete the work plan (more significant for larger requests). It takes work to burn. Sufficient merit-reviewed funding (if you have it). Staff, both number and experience.
        • Local computing environments (why aren’t your local resources sufficient?) 3:1 demand request vs. what’s available
        • Special needs (ex: weather projects may be seasonal)
        • Other access to HPC resources – campus, Blue Waters, etc.
      • Other docs – try to limit to 2-3 pages
        • CVs for PIs and Co-PIs
        • Publications in the XRAS submission form (NSF and service providers see this as accomplishments – it’s important to publish)
        • Optional documents if you need more space to justify something in your main document (only if needed!)
    •  Proposal Review Criteria

      • Methodology
      • State Appropriateness of Computations for Scientific Simulations – step size, time scale, etc.
      • Describe Efficiency in Usage of Resources
      • Computational Research Plan

      XSEDE Project: Like a “bank account” for allocations. Permanent, only one per PI. Can transfer, supplement, extend, but it’s typically thought of as a 1-year allocation. Don’t use them, you lose them.

      Eligibility: PI must be researcher or educator (or postdoc) at U.S.-based institution

      Overview of Research Request:

      Portal account at xsede.org
      Portal.xsede.org -> Allocations -> Submit/Review request
      Web forms
      Pdf upload for proposal and CV
      ~ 2.5 months b/w deadline and award
      Deadlines: 15 Oct, Jan, Apr, July

   

 

  •