Distributed CSE Support
How to apply
To help validate and discuss the rationale for the work being proposed it is recommended that applicants contact the NAG CSE team at [Email address deleted] at least 2 weeks before the application deadline, as this will allow time to discuss the dCSE proposal in further detail. The dCSE review panel expects a thorough performance analysis of the codes in question as a basic justification of a dCSE proposal, and the NAG CSE team can contribute to ensure that these requests are met as well as advise prospective applicants on related work or the typical scope within the dCSE framework.
The Proposal and Coversheet
For the main submission please fill in the coversheet which is to be submitted along with the main part of the application. The main part should be at most 8 sides A4 or 3800 words in length and include the following information:
1. Overview of project (1 side maximum)
This should describe the scientific project in general terms. Where the applicant is an existing HECToR user they should provide details of the grant(s) involved and, their (expected) start and end dates, and progress made to date. If the applicant expects to apply for additional AUs to carry out the CSE project this fact should be clearly indicated and a HECToR Application form completed for a Class 2b request of resource.
2. Applicants' track record (1 side maximum)
This should describe the applicant's track record in the fields of software engineering and HPC, not the general science background. In particular, if the applicant has contributed to the development of HPC software in the past, then he or she should illustrate the development and platform history of the code, i.e. major algorithmic changes, functional additions or rewrites, the list of hardware and software environments the code is running on, and the performance of the code on these environments.
3. CSE support requested
This should detail the programme of software development work in terms of specific and quantifiable milestones as well as the personnel resources (in person months) required. It should describe the code(s) to be worked on and (if known) the approach to be taken and comment on the goals, in terms of scalability and performance advances expected, and the anticipated problems and possible solutions.
Graphical representations of workplans, dependencies and key milestones are encouraged.
If the applicants would like to suggest one or more particular individuals as candidates to supply the support then they should name them and attach a brief CV describing their expertise in HPC. Otherwise they should outline the skills that they feel are necessary to do the work.
4. Computational and Scientific benefits (1 side maximum)
This section should identify how the work described in the previous section will impact the applicants' scientific output, justify the need for such a project w.r.t. to the science community needs and assess its impact on both a national and international level with work being undertaken by other groups in the area. Plans to share code modifications and experience gained within the scientific and computational communities should be included here. Where there is a clearly identifiable central source for a code applicants are required to provide assurance that their proposed code modifications will become incorporated. In such cases a letter of support from the code keeper outlining the intent to include and publicise any or all proposed code modifications in the main source should be provided.
Applicants are strongly advised to give specific examples of new scientific research that will be facilitated by their proposed work. This may be from their own research plan or from a direct sponsor or group of sponsors requiring the code improvements. There should be a clear explanation of the proposed scientific cases that are enabled by the completion of this work. Letters of support from UK research groups which give details of their scientific justification and HECToR requirement for the code development are also encouraged and this is especially relevant for established codes. Such letters may be sent accompanying the main submission. Where there are only a limited number of immediate users for the code improvements there should be a clear demonstration for the future applicability of a code (e.g. replacing a current code, enhancing future capability, attracting a new user group to HECToR).
5. Background Information (1 side maximum)
As part of the proposal applicants should include the following information:
- Briefly summarise current use of the code, both on HECToR and other HPC systems, by themselves and others. In particular describe systems used, numbers of processes, problems sizes and magnitude of runtimes etc.
- Where possible include measurements of current code performance, in terms of computational performance and/or scaling.
- If specific software, algorithms, or other materials are to be used within the project key references describing these should be provided.
Points to consider for the main proposal
Applicants are recommended to consider the whether the following questions are addressed in their proposal.
- What is the scientific outcome of the work?
- How do the aims of the work achieve better use and efficiency of the code?
- Are the proposed developments likely to bring about increased usage of the code on HECToR?
- Will there be any knowledge gained from the work which may be transferable to other applications?
- Might the proposed work bring about new scientific exposure for HECToR?
- For established codes will the proposed work be released to a wide user base or be introduced to the main code release?
- Does the likely scientific outcome enabled by this work on HECToR warrant the level of effort proposed?
Requested Resources / Work Plan:
- Can the proposed work plan be completed within the requested amount of time?
- Are the personnel resources requested adequate?
- Is the approach a sensible way to go about the proposed developments?
- How does the work plan implement the approach?
- Is the time frame adequate?
- Are the resources requested sensible?
- How do similar codes approach the problem?
- Is the proposed method of achieving the scientific aims the best way of approaching the work? (Are there other codes or techniques that might achieve the scientific aims more efficiently?)
Applicants and Proposed Personnel:
- Is the track record of the applicant in software development and management explained?
- Is there a proposed candidate who is suitable for doing the work? If so, will the work be initiated within a reasonable amount of time?
Further help and information
Please see the dCSE FAQ section for help, or go back to the main dCSE page. For further assistance, the HECToR CSE support team can be contacted at [Email address deleted], or via the HECToR Help Desk.