Advanced Scientific Computing Support Service
This document provides guidelines on how to use the Advanced Scientific Computing Support Service of the School of Physics at Sydney University.
The Advanced Scientific Computing Support Service facilitates research within the School of Physics by providing advice and support for advanced scientific computing. The service provides advice on advanced techniques, such as numerical analysis, parallelization, user interfaces, visualization, data transfer and data analysis. The service also helps the School of Physics to identify, maintain, and build skills and assets. For example, the service will provide the tools to build and maintain data repositories and repositories of reusable scientific software.
Why would you consider using the service?
The service offers advice and support for advanced scientific computation, as opposed to general purpose software development or hardware or systems support. For hardware and systems support, read the FAQ and use the Physics IT Support Request Form.
While most of the skill base in computational physics already resides within the School, your research group may be facing an advanced computational physics task where there is not enough expertise available to complete the task. Relevant tasks include making code run more efficiently on parallel computers, data reduction and visualization of data sets. The service gives advice and support to help your group to perform the task for itself. This increases the level of skill within your own group. The service also helps your group by providing tools to maintain reusable archive copies of the software produced in the course of its research.
Before you consider using the service
Before you consider using the service, identify the advanced computational physics task and take a look at your research group. What is the skill level and availability within your group? Could your group complete the task by itself? If not, why not? Is your project at a point where using the service could help?
There are a number of other considerations to take into account, depending on the type of task involved and the type of service your group requires. The following list is being progressively updated as more services become available.
- Data transfer
This service includes the facilitation of novel and advanced data transfer, as opposed to routine data transfer. To be able to use this service your group needs to ensure that the data transfer requirements are well documented and that certain facilities are available as per a data transfer checklist. - Software archiving and reuse
This service will soon offer software tools and an an archive structure to enable your group to archive reusable software. To be able to reuse, modify or archive an existing software package, you need to ensure that the software passes a reusability checklist. - Parallelization
(Not yet offered.) Parallelization involves a number of issues in addition to the issues involved in software reuse. - Numerical analysis; User interfaces; Advanced visualization
Not yet offered.
Priorities
Your request will be given a priority according to how well it fits within the scope of services offered. Requests which fit poorly will be given a lower priority. Requests which are considered to be out of scope will be politely declined.
Training, reading and resources on the web
Further training or reading may increase the level of advanced scientific computing skill in your group. Postgraduate researchers may want to consider enrolling in or auditing one or more of the advanced Computational Science courses offered by the university, especially COSC 3912, Parallel Computing & Visualisation. Some computational science resources available via the web, including courses, organizations and software are listed in a separate document.
Contact Details
Contact Paul Leopardi via email to leopardi@physics.usyd.edu.au or call extension 16081.