Workshop Goals - PowerPoint PPT Presentation

About This Presentation
Title:

Workshop Goals

Description:

Presentations of application-science needs, worries and frustrations ... In other words, avoid commercials' for particular CS approaches to solving a problem ... – PowerPoint PPT presentation

Number of Views:24
Avg rating:3.0/5.0
Slides: 8
Provided by: richard1003
Category:

less

Transcript and Presenter's Notes

Title: Workshop Goals


1
Workshop Goals
  • Richard P. Mount
  • May 24, 2004
  • DOE Office of Science Data Management Workshop

2
Workshop Progress (1)
  • SLAC DM Workshop, March 2004
  • Presentations of application-science needs,
    worries and frustrations
  • Presentations of relevant computer science and
    technologies
  • Gap between many application sciences and CS
  • We need a holistic approach to our scientific
    workflow problems and these CS guys are talking
    about ontologies!
  • Gap between what CS is funded/rewarded for
    producing and what the application sciences need
  • Proposals for long-term user support and product
    hardening do not get funded and the work wouldnt
    get you tenure.
  • Much vigorous and productive discussion

3
Workshop Progress (2)
  • Extended Organizing Committee, April 2004
  • Application scientists and computer scientists
    searching for, and finding, mutually
    understandable ways to organize the issues
  • Workflow diagrams proposed as a good basis for
    comparing and contrasting application science
    needs (in many cases)
  • Agreement on a straw-man report structure to be
    presented to this workshop

4
Chicago Workshop Goals
  • Discuss and improve the proposed report structure
  • Continue the move from a laundry list of needs
    and topics to well-organized programs of work
    reflecting application-science priorities
  • Workflow diagrams may help in this organization
    of ideas
  • Aim to make a case at the Program of Funding
    level, the WBS level can be used for illustration
  • In other words, avoid commercials for
    particular CS approaches to solving a problem
  • Application-science priority means we would be
    prepared to contribute to this work with our
    BES/BER/FES/HEP/NP
  • Propose approaches to funding and program
    management that exploit this mechanism (e.g.
    SciDAC?)
  • Priorities will change the approach should
    still work
  • Address the gap between academic CS and the need
    for robust software with 20-year support
  • Funding and program management mechanisms that
    support what is needed

5
Workshop Structure
Monday May 24 Monday May 24
0915 am Straw-man report
0945 am View from simulation-driven applications
1045 am View from experiment/observation-driven applications
1115 am View from information-intensive applications
1145 am DOD data-management requirements
130 pm Technology Working Groups Groups to address a) additions, corrections and refinements to the report, and b) gaps, cost, priorities and classification into development/hardening/deployment categories. Groups 1. Workflow, dataflow, data transformation 2. Storage, data movement, grid, networks 3. Metadata management and cataloging 4. Efficient access and query, data integration 5. Integrated data Analysis, visualization
430 pm Panel Discussion on Workflows Make progress towards a common approach to describing workflows such that commonalities and true differences are clear.
6
Workshop Structure
Tuesday May 25 Tuesday May 25
0830 am Group Leads will report back to the general session, summarizing their group discussion. This will be followed by open discussion. This should lead to cross-group comparison of cost and priorities.
1100 am Panel on a management plan - an effective process to direct funding to benefit the various application domains. This is the plan that should get DOE application offices (other than OSCAR) interested and involved.
115 pm Parallel sessions of application-led groups to prioritize technologies in their domains. Groups simulation-driven applications experiment/observation-driven applications information-intensive applications
345 pm Brief summary reports from application-led groups. Followed by a panel session to come up with a prioritization and cost for the development/hardening/deployment of SDM technologies.
730 pm Group leaders and other volunteers will develop a joint plan on how to normalize the cost and priorities.

7
Workshop Structure
Wednesday May 26 Wednesday May 26
0830 am Present the cross-cutting matrix and priorities/cost to the general session
900 am Open discussion led by panelists
1100 am Assignment of coordinators for final writing.
1200 pm Adjourn
Write a Comment
User Comments (0)
About PowerShow.com