Climate Simulation Application for Computational Resources

Request for Climate Simulation Laboratory Computational Resources

Note: CSL proposal deadline is December 16, 2005.

Allocations will begin February 1, 2006 and end July 31, 2007.

Instructions: This web-based form constitutes an official request for computing services available through the Scientific Computing Division of the National Center for Atmospheric Research. Please use the documents on the CSL Home Page as guidelines to help you complete your request.

There are two options for submission:
(1)Continue if you want to use this form for Sections A,B,C,D,E and G. Section F must be submitted as a PDF and section H as a second (separate) PDF.
(2)Click here to download a Word form for email submission to alloc@ucar.edu for Sections A,B,C,D,E and G. Section F must be submitted as a PDF and section H as a second (separate) PDF.

Please fill in all of the fields on the form. Review the information you provided in this section. You may, at any time, correct or change the information you've provided but, you must do so before submitting the form. Once you have submitted the form it is not possible for you to go back into this form to make changes. Once you have verified the information you provided, click on the "SUBMIT Request" button at the bottom of the page. You may send edits to alloc@ucar.edu All changes and any new pdf files are due by December 16, 2005 at midnight.

You should get a confirmation within a few days. If you do not or if you have questions on the CSL allocation process contact alloc@ucar.edu call 303-497-1229.

A. Project Information

A1. Project title or project number if this is an extension request:

A2. State the TOTAL NUMBER OF GAUs needed from February 1, 2005 through July 31, 2007. Assume all use is in the REGULAR Queue. If your need for resources is different for different months, then please list the GAUs needed for each of the 16 months. If you have deadlines relevant to your CSL research that you have to meet, please give a brief discussion of them and how they affect your computing requirements.

B. Identification of the Primary Principal Investigator

If a PI will also be a programmer on the project, type "+PROGRAMMER" after their name, and it will not be necessary to duplicate their information in the Section E, "Identification of Programmers".
B1. PI's name:

B2. Title of PI (Professor, Research Associate, etc):

B3. Department/institution:

B4. FedEx Mailing Address (FedEx will not accept a PO Box, campus box, or a drop box.):

B5. Email:

B6. Telephone:

B7. FAX number:

B8. Brief summary of education of primary principal investigator:

B9. Point-of-contact for the CSL project's computational aspects:
For each CSL project, SCD uses a point-of-contact. This may be the Principal Investigator, the Requesting Scientist, or a programmer on the project. Notices of machine changes, resource allocation changes, and machine utilization will be directed to this person. We also solicit input on proposed changes and their potential impact. Please include telephone, email and postal mailing address for the point of contact:

C. CO-PI(s)

Use the text box below to identify all CO-PI(s) on the project. If a CO-PI will also be a programmer on the project, type "+PROGRAMMER" after their name, and it will not be necessary to duplicate their information in Section E, "Identification of Programmers".

Include the following six items for each co-principal investigator:

CO-PI's first and last name
CO-PI's title
CO-PI's institution and department
CO-PI's telephone number
CO-PI's fax number
CO-PI's email address
CO-PI's FedEx Mailing Address (FedEx will not accept a PO Box, campus box, or a drop box.)

D. RESEARCH SPONSOR(s)

Please provide the following information for all sponsors of the project.

D1. Name of research sponsor:

D2. Mailing address of sponsor:

D3. Program name:

D4. Agency director/contract monitor and email address:

D5. Award number (or proposal number if pending):

D6. Start date of award:

D7. Date of award expiration:

D8. Principal investigator for the award:

D9. PI's complete mailing address if not given in Section C:

D10. PI telephone if not given in Section B:

D11. Email address of PI if not given in Section B:

D12. Use the text box below to identify additional sponsors.
Please list each additional sponsor and provide the information for each sponsor as you did for the sponsor previously identified in C1-C11. All 11 items are required for each sponsor.

E. Identification of Programmers

Use the text box below to identify programmers or assistants who will be working on this project. It is not necessary to list information for PIs or Co-PIs who want computer accounts if you noted their names with "+PROGRAMMER" in Section B or C.

Please include the following four items for each programmer listed.

Programmers first and last name
Telephone number
Fax number
Email address
FedEx Mailing Address

F. Computational Requirements

Email Section F to alloc@ucar.edu as a pdf file.

Details on how computing requirements are estimated MUST be included. This is because reviewers of the CSL computing requests are asked to evaluate whether the resources sought are justified and will be used efficiently.
Information for items 1 thru 9 is required. Please number each of the 9 required items in your response.

Number of GAUs required per run. This should be substantiated with benchmarks on an IBM Power3 or Power4 system. You may request an account to obtain this information by completing theRequest for Computing Services Form.
Provide information for 2, 3, and 4 for each model you plan to run.
CPU time required per run.
Node-reserved time per run, which is defined as number of nodes * number of processors per node * wall clock time for a run. Please specify each variable used in this equation.
Memory requirements per node.
Number of runs required for each numerical experiment. State why this number of runs is needed and not fewer or more runs. Without this information the Panel may reduce or deny your computing request.
Evidence that your program code is well optimized and operates efficiently and responsibly in an environment where resources are shared and are at a premium. Include documentation on program code performance (e.g., timings, performance monitoring tools). The PI is responsible for convincing the reviewers of optimized code performance.
Plan for the analysis of model results. Note: CSL users have access to separate data analysis computers. See CSL Equipment Please address item #3 under Allocation Process in the CSL Announcement of Opportunity.
A. Special requirements, such as large memory, large disk space, special software, or long run times without restarting.
B. What is the maximum number of GB you plan to store on NCAR's Mass Store System (MSS) during the life of your project? If your data storage is < 500GB just enter < 500GB. This information will be used to estimate future MSS requirements.
Discuss your plan for sharing the data generated by your CSL research with the community including type of data to be shared, groups with whom you plan to share your data, length of time your data needs to reside on SCD's MSS after your CSL project ends to meet these needs. This discussion should address item #4 under Allocation Process in the CSL Announcement of Opportunity.

G. Publications

Please list publications relevant to the science and computational aspects of this proposal. If you list more than five publications, highlight with an asterisk those publications which you want the panel to especially note.

H. Proposal Summary

Email Section H as a PDF file to alloc@ucar.edu
(Section H must be a separate PDF from the PDF sent for Section F.)
Please provide a self-contained proposal describing your project. This should be the equivalent of 4-5 typewritten pages (with a maximum of 5 pages), excluding figures and publications. It is important to give sufficient information for the realistic evaluation of the problem.
The description of the numerical approach should be sufficiently detailed for reviewers to confirm the estimate of machine time required. For example, in the solution of a system of time-dependent partial differential equations, the descriptions should include the finite difference approximations, and the methods proposed for integration in time and space. An estimate of the number of grid points and the number of parameters to be retained at each time step should also be given.
The CSL Allocation Panel will make recommendations for CSL use based on the four CSL criteria and the allocation process both listed in Announcement of Opportunity. Please label your responses to the criteria as Criteria 1-4. Include sufficient documentation for the panel to easily determine if the proposal meets the CSL criteria and other factors under Allocation Process.
Insufficient documentation will mean a denial of your proposal or a reduction in the resources allocated, until the next CSL Announcement of Opportunity.

Please fill in all of the fields on this form and, when finished, review your information. Click on the "SUBMIT Request" button at the bottom of the page. Please use the documents on the CSL Home Page as guidelines to help you complete your request.

You may, at any time, correct or change the information you've provided but, you must do so before submitting this form. Once you have submitted this form it is not possible for you to go back into this form to make changes. You may however, send edits to Database Services alloc@ucar.edu All changes and any new pdf files are due by December 16, 2005 at midnight.

IMPORTANT: If you submit a request for computing services you should get a confirmation within a few days. If you do not, please send email to alloc@ucar.edu or call 303-497-1229

Please contact alloc@ucar.edu concerning any questions on the CSL allocation process, or for questions about completing this form.

Before you submit this request, please review all entries by clicking on the button below.
....

(Updated 10/21/05)
Questions? Send email to: Ginger Caldwell, alloc@ucar.edu or 303-497-1229
Broken links? Send email to: Michelle Smart, msmart@ucar.edu