1 / 8

Possible Strategic Capability Projects

Possible Strategic Capability Projects. High resolution coupled simulation: (ASD) 1/8 o CAM, 1/10 o POP, 20 years (initially) SCIDAC proposal: Small, Bryan, Tribbia , Dennis, Park, Tomas, Niklas Schneider, Young-Oh Kwon, Saravannan Coupled intra-seasonal (seasonal) forecasting

salali
Download Presentation

Possible Strategic Capability Projects

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. Possible Strategic Capability Projects • High resolution coupled simulation: (ASD) • 1/8o CAM, 1/10o POP, 20 years (initially) • SCIDAC proposal: Small, Bryan, Tribbia, Dennis, Park, Tomas, Niklas Schneider, Young-Oh Kwon, Saravannan • Coupled intra-seasonal (seasonal) forecasting • 0.25 CAM nudged to analyses; • focus on MJO and it’s 1-2 week impact on western US • Rich Neale and Julie Bacmeister • High resolution WACCM run: • 1/8o, 0.5-1km vertical resolution (150-300 levels); • 1 year runs for 3 cases (solar max, min, medium) • Hanli Liu, WACCM group • Simulations to produce a fully coupled data assimilation product for 1960-2007 • 1-degree CESM • GokhanDanabasoglu, Joe Tribbia, Alicia Karspeck, DART group

  2. CSL Proposal • CSL = 28% of Yellowstone (~250 million core-hours over the 18-month allocation) • For current CSL, CESM has ~68% of resources • If keep same target, then ~175 million core-hours for CESM. About 20X the current CSL(?). • Possibly argues for some big, high-level, across-WG projects • Large Ensemble Integrations?, • High(er)-Resolution Controls?, • Long Control Integrations with CESM1.5? • Additional resources for WG development/production efforts

  3. CSL Proposal Current Resource distribution across working groups Could keep same percentages - acknowledging CV and CC merger, SDWG Modify based on use or community needs/interest?

  4. CSL Proposal Possible Large Across-WG Projects? • Large Ensemble Example • CESM(CAM5) 1-deg costs ~1930 GAU/year with a data volume of 26.4 GB/year • 1 core-hour on Yellowstone equivalent to 2.15 GAUs (?) • CESM(CAM5) 1deg = ~900 core-hours/year (?) • 50 ensemble members from 1900-2099 equals ~9 million core-hours (~5% of request) • Could also perform ensemble runs with WACCM, FASTCHEM, BGC for comparison? • (what if ~20% of total request committed to across WG projects? Still have ~15Xs current CSL allocation)

  5. CSL Proposal • CESM1.5 control runs • CAM5 – 1deg = 1930 GAU/yr (900 core-hour/year) • Assume computational increase (1000 core-hour/yr) • Assume 1000 year run • ~1 million core-hours

  6. CSL Proposal • For NWSC, proposals due Feb 15, 2012 • Proposed Timeline: • January 6: Drafts of Working group reports and past accomplishments • undergoes cross-referencing and refinement • January 13: Final WG drafts available • January 25: SSC proposal draft due • Comments from WGs back by Feb 3 • Feb 8: Final draft available • Feb 14: Proposal submitted

  7. Justifying resource needs • HPC — similar to current practice • Cost of runs necessary to carry out experiment, supported by benchmark runs or published data • DAV — will be allocated, similar to HPC practice • A “small” allocation will be granted upon request • Allocation review to focus on larger needs associated with batch use • Memory and GPU charging to be considered • HPSS — focus on storage needs above a threshold • 20-TB default threshold initially • Perhaps lower default for “small” allocations” • CISL to evaluate threshold regularly to balance requester/reviewer burden with demand on resources • Simplified request/charging formula • GLADE — project (long-term) spaces will be reviewed and allocated • scratch, user spaces not allocated (Courtesy of David Hart, CISL)

  8. Working Group Template • Research Plan and Broad Overview Objectives (1page) • Development Objectives (1 page) • Coupling across components and understanding interactions • New parameterizations/processes • High-resolution and new dynamical cores • Addressing biases/shortcomings • Software development • Production Objectives (1 page) • Coordinate modeling and assessment activities • Benchmark simulations • Climate variability and predictability • Proposed Experiments and Computational Requirements • Development Experiments (with Table) • Production Experiments (with Table) Table Example

More Related