On 16 November 2004, NUS and Singapore Computer Systems (SCS) launched an initiative to build the Tera-Scale Campus Grid (TCG@NUS), a world-class supercomputing grid at NUS. TCG@NUS is Singapore's first large-scale campus-wide grid implementation which harnesses idle computing capacity to generate high-speed computing power. TCG@NUS is a groundbreaking regional project, as it is the largest of its kind among educational institutions outside of the US and Europe. Initially, 800 computers will be aggregated to form the grid. SVU is proud and encouraged to be given the opportunity to drive this initiative and grateful for the contributions and support from Faculty of Engineering, Faculty of Science (Physics and Mathematics), Business School, Central Library and School of Computing. With the addition of this powerful PC Grid, SVU is better equipped to support a wide range of parallel research computing applications in NUS. We have the shared-memory multi-processor system to support large multi-threaded and OpenMP parallel applications. Our HPC cluster supports Message Passing Interface (MPI) based distributed-memory applications. And this PC Grid will facilitate large scale data parallel and parametric study applications. HPC technologies and applications will evolve with time. To cater to the needs of our users, SVU will monitor closely the trends and developments in HPC technologies. Areas that we will explore in the future include the emerging Grid technology, high-performance file system, storage and cluster interconnect. We will keep you informed of developments and enhancements to our resources, activities and capabilities. In this issue of our newsletter, we would like to share with you some articles on Grid computing, applications software and IT security.

Happy reading and Best Wishes for the New Year!

The recently announced Tera-scale Campus Grid at NUS (TCG@NUS) initiative will provide an unprecedented amount of computing resources to support research computing in NUS. TCG@NUS is an initiative to create new research computing capability at NUS by harnessing the idle compute cycles of PCs across campus. The campus-wide Grid will be implemented using the United Devices GridMP Platform software.

It is inevitable that comparison will be made between the PC Grid and the dedicated HPC cluster. Even though both are based on the concept of aggregating the compute power of large number of computers to solve a large scale problem, the target applications are quite different on both systems. In a nutshell, PC Grid is more suitable for running large scale data parallel and parametric study applications whereas the dedicated HPC cluster is more suitable for running tightly coupled parallel applications that require low latency interconnect. Please click here for more details of the comparison. You may also check out some FAQs on the TCG@NUS at SVU website.

While, most of SVU's systems have strong security measures, users also need to be aware of the part they can play in making SVU a more secure environment for everyone. Userids and passwords need to be managed properly, including regular changing of password, choosing strong passwords and using secure means of logins to SVU systems. With everyone's cooperation in observing these simple guidelines, we'll be able to maintain a safe and secure environment for everyone to work in. Remember, if a system is compromised, everyone suffers. Please click here to read the full article to learn the details and feel free to write to should you have any questions.

SVU is pleased to announce that one of its users, Mr Lim Wee Chuan, Eldin, won the second prize (under the Institutes of Higher Learning category) in a recently concluded national competition in Grid Computing. The SG@home Ideas Competition sought to elicit ideas from all levels of the educational spectrum and the general public for a large grand challenge; project using Grids/Grid Computing. Photos of the prize presentation ceremony may be found at this website. Read Wee Chuan's account of his experience and his submission entitled Turbulence Modelling on a Grid Computing Platform.

Parallel computing of Fluent is widely used by researchers in NUS. With the help of parallel computing resources, users can shorten the time to solution significantly. However, most of CFD simulations take very long time to converge even with the advantages of running jobs on multiple CPUs. Computing time up to days, weeks or even months is not uncommon for CFD users. However in a multi-user shared environment, each job is given a limited time to run. Usually the long running job will be terminated automatically by the job scheduler when the given time is up. Therefore it is necessary to know how to backup the intermediate results so that the terminated job can be resumed using those intermediate results rather than having to restart from the beginning. Please click here to learn how to do such backups in Fluent, and email should you have any questions.

The use of the advanced information technology in the study of biological problems has been of great help in facilitating the understanding of biological processes. Scientists apply the principles of information sciences and technologies to make the vast, diverse and complex life sciences data more understandable and useful. Today, many such data and tools are available on-line for free downloading. A brief summary of these open source tools (extracted from a tutorial in SC 2004) is given in this article as a reference for researchers who are interested in this area. read on ...