- Welcome new members
- 4 new members: Lisa Arafune, Ken Bloom, Rudy Eigenmann, Ani Ural
- Thanks to all members for your time and efforts.
- John happy to do an orientation for any new members. Lisa, Ani, Ken, Rudy are all interested in doing this. Any other members who would like to join are welcome.
- Plan for previous summaries
- Acknowledgment of 2019 XAB meeting schedule
- 2019 Feb Call: Wed, 2/6 – 12-1pm ET | 11am-12pm CT | 10-11am MT | 9-10am PT
- 2019 Apr F2F Mtg: Tue, 4/16-Wed, 4/17 – Dinner the night before and 8am-4pm CT meeting in Chicago area
- 2019 June Call: Wed, 6/26 – 12-1pm ET | 11am-12pm CT | 10-11am MT | 9-10am PT
- 2019 Aug Call: Wed, 8/21 – 12-1pm ET | 11am-12pm CT | 10-11am MT | 9-10am PT
- 2019 Oct Webinar: Wed, 10/9 – 12-4pm ET | 11am-3pm CT | 10am-2pm MT | 9am-1pm PT
- 2019 Dec Call: Wed, 12/11 – 12-1pm ET | 11am-12pm CT | 10-11am MT | 9-10am PT
| Karin|| Calendar appointments sent out for all 2019 meetings|
Summary of John's visit with NSF on 27-Nov
- John was invited by OAC office director Manish Parashar.
- Considering what a follow-on solicitation post XSEDE 2.0. Last time John had been invited was 2011.
- Helpful for them to hear point of view from PI
- Presentation included:
- Questions: Is there an NSF vision for what is desired? Is this a step in a long-term exit strategy for NSF? What are the budgetary constraints? What strategic impacts are most important to NSF?
- They had thought about some of these things but didn't have many answers.
- Offered various suggestions:
- Go Big (expanded scope working to reduce duplication across the foundation)
- Go Really Big (single MREFC-level award; treat CI as a facility)
- Go Big and Wide (multiple awards that address various service areas)
- Go Home (fund ramp-down; SPs collectively manage ecosystem; encourage creation of commercial spin-off)
- Status Quo (fund similar project at similar budget)
- Keep the trains running (fund something like XSEDE at significantly reduced budget)
- Whatever follows XSEDE 2 needs to satisfy needs of community & avoid disruption.
- John's personal interests are in building infrastructure in support of research. Can do operational side but not as interesting. Keep the trains is not as interesting as a "Go really big" option. Leadership changes can bring new perspectives.
- John's presentation did not include Power Wireless, Earth Cube: Funded program office that runs competitions. That model wouldn't work in our timeline. Would take more than time we have available.
- Karen: hired as program office mgr with Neon in 2007 or so. That model is complicated. Had to keep renewing every couple months. No time to do any work. Hope we don't go down that path with XSEDE. Doesn't work.
- Ken: In past year + open science grid had similar experience. Coming to end of funding, sustaining infrastructure is not NSF's business. Had to figure out what new OSG was–smaller grants to different domain areas being used to fund joint project. How that project gets managed. Conversation with Frank Würthwein would be valuable. John has talked with him.
- John anticipates being invited back, so any alternatives/suggestions are welcome
- Spoke with program officers from elsewhere at NSF in afternoon. Provide info on why they should be interested in XSEDE. Pie chart that shows to whom we deliver allocated resources by their funding agency. XSEDE mandated to support researchers funded by any federal agency. NSF funded PIs rec'd about 41% of delivered resources. DOE 14%, NIH 14%, DOD 5%, NASA 4%, DOC 3%, None 11%; All others 8%. Supporting broad range
- Distribution of consumption by NSF-supported research projects: math & physical sciences 48%/$12.8M/3000 node cluster; bio 24%; eng 15%; geo 8%. NSF projects consumed ~7.9B core hours.
- Working to deliver to each directorate AD a report for last fiscal year with complete list of funded PIs with allocations, $ value of awards, $ values of cycles consumed, resource dedicated to them. Allow them to connect funded PIs to services provided by XSEDE.
- Albert: is that info in public domain? Out there but hard to get at. Compile on website that is easily accessible? XD Mod collects all usage/allocation information & provides portal to look at the info. All of this data is there except for some of the value info. Not easy to extract from XD Mod, so this will package in a way that is easier for directorates to view.
- Info used by MPS to advocate for more resources from XSEDE follow-ons? Don't know how they'll use it but want to enlighten them as to what is actually being done by their funded PIs & engage in discussion with OAC to protect what they are already getting from XSEDE. May have science directorates contribute to the program. Trying to prompt conversations that will help directorates and help define better program going forward.
- Karen: We should try to provide guidance to NSF. Hard to rely on people at NSF for national guidance. Tell them what our adv board thinks instead of try to reverse engineer to get funding.
- Rudy: NSF is community driven and we are part of the community. Did possibility come up that you would initiate community workshops that would help inform future direction?
- JT: They feel they've addressed through RFI workshops etc. A lot of these issues came up. Reaction to inputs by ACI. Work hard to align our activities with NSF strategic plan that covers 2018-2022. Until last year had 2 other documents that guided us which have expired. Are there going to be follow-on docs that help inform community?
- Karen: Dept. of Energy adv board. Does XSEDE have a relationship with Oscar/DOE? Keep communications open but nothing deliberate. DOE relies more on what's in the wind/what they hear. Missing component about connecting pieces. Important to make connections when you can.
John to provide slides