E-Research Case Studies at VUW

Astrophysics: grid computing helps VUW researchers find new planets

microlensing.gif

Denis Sullivan and his PhD student Aarno Korpela used grid computing to help them detect new planets: they estimated that half a year's modelling could be done in a day, using the grid of around 200 UNIX workstations based in Maths, Statistics and Computer Science.

Gravitational microlensing - a process of measuring the lightcurves from stars - extends the planet detection range and is important in estimating the number of planets, and ultimately, the likelihood of life in other solar systems. In March 2005, Aarno discovered a planetary system located in the next inner arm of the Milky Way, some 17,000 light-years away.

Aarno and Denis made an even more exciting discovery in July 2005: a rocky or icy planet about 23,000 light years away. OGLE-390 is Earth-like in size, and although the planet is not Earth-like in its conditions for life, this discovery was one step towards the goal of finding a true Earth-like extra-solar planet. The grid made a significant difference to the speed of this research. A typical modelling task might involve 50 different planet masses and 50 different planet distances, giving a total of 2500 processing jobs. With each job lasting about two hours, the task would take 5000 hours (more than half a year) on a single workstation, but with the grid could be completed in around 28 hours.

Denis has future plans to extend his grid use to the modelling of pulsating white dwarf stars.

Pacific Studies: uniting NZ Pacific postgraduates using the Access Grid

The Pacific Postgraduate Talanoa Network connects sites in New Zealand for interactive videoconferencing and collaboration sessions using the Access Grid via the Kiwi Advanced Research and Education Network.

The Talanoa Network was initiated and co-ordinated by Va'aoman%u016B Pasifika, Victoria's Pacific and Samoa Studies unit, and is the first resource of its type for Pacific students.

Associate Professor Peggy Fairbairn-Dunlop, Director of Va'aoman%u016B Pasifika, says running the Access Grid via KAREN is of immense value to Pacific researchers.

"The Access Grid sessions provide students with a place to both present their research and raise any issues. We have some very lively debates on issues relating to Pacific research methodologies.

"The Network also helps to break down feelings of isolation which many of our small but growing group of Pacific post graduate students may feel, scattered as they are through New Zealand."

The sessions have grown in popularity since a launch in June 2007, with an average of 25-30 people attending the fortnightly online seminar sessions. Tertiary Education Commission and Ministry of Education staff are among the participants and the University of the South Pacific and the National University of Samoa are keen to link to the seminars.

Nanotechnologies: high performance computing with the BlueFern accelerates simulations

nanoparticle.bmp

PhD student Dmitri Schebarchov is one of several VUW staff and students using the new BlueFern supercomputer to accelerate their research into nanotechnologies.

Dmitri's research into thermal and mechanical properties of metal nanoparticles would be impossible without the BlueFern's 4000+ processors and extremely fast communication between processors. He says: "The time taken for a single caloric curve simulation has reduced from several months to a couple of days. Hundreds of simulations (at different temperatures) can be run at the same time, and each simulation can be spread over multiple processors to speed things up even more."

Internet security: grid computing for statistical analysis

Computer Science PhD student Christian Seifert is investigating the use of client honeypots; security devices for finding malicious servers on the web ( e.g. a web server that attacks a web browser). Christian used the MSCS UNIX grid to statistically analyse a large quantity of web pages for malicious content. The analysis would have taken many weeks to complete on a single PC, but Christian reduced this to a few days by splitting his analysis into chunks and submit these jobs to the grid.

Heritage digitisation: grid computing for large website generation

In 2007, the New Zealand Electronic Text Centre investigated the feasibility of using VUW's computing grid as a platform for website generation.

In 2006 the NZETC had worked with the National Library of New Zealand to digitise more than 60,000 pages of the Transactions and Proceedings of the Royal Society of New Zealand. Generating the website to deliver this material online was very time-consuming, requiring at least 8-10 days of processing. The grid-based application developed by the NZETC proved the grid was a good platform for generating large websites from XML source materials. The application was able to generate around 80,000 web pages in about 4 hours, making it many times quicker than a system based on a single computer.

Stability and evolution: grid computing for ecological simulations

Richard Mansfield's PhD research involved investigating several properties of artificial 'rock-paper-scissors' ecosystems, in which each of three species can only invade the next species in the cycle. By running cellular automata simulations, and counting the number of generations until an extinction occurs, Richard showed the effects of population size on increased stability.

Richard obtained many of his results using computer simulations run on the grid. The grid was especially useful for this work because most of the problems could be split into small tasks.

BRCSS: building social science research capability using the Access Grid

vuw-ag.JPG

Social scientists at universities around New Zealand have been collaborating since 2005 using the Access Grid. The BRCSS Network brings together a platform of externally funded programmes, and aims to contribute to the development of new and emerging researchers, to build new collaborative e-research teams, and to extend local and international linkages.

BRCSS researchers use the Access Grid for a range of activities including management meetings, seminars and workshops, and postgraduate research seminars.The Acess Grid enables discussion of ideas across organisations and access to international experts, who can attend virtual discussions and seminars regardless of where they are located.

Music: Geophony transforms real-time earthquake data into music

New Zealand School of Music researcher Dugal McKinnon, has collaborated with GNS Science to produce a sound installation, Geophony.

Geophony is a multichannel sound installation based on sonification of multiple real-time seismic data streams provided by GNS Science via the Kiwi Advanced Research and Education Network (KAREN). The installation was the first project at VUW (and possibly in New Zealand) to consciously use KAREN for a project in the arts, and specifically in music.

The project builds on VUW's track record for interdisciplinary collaboration between scientists and artists, including the successful Are Angels OK? initiative between VUW's International Institute of Modern Letters, the physicists of New Zealand, and the Royal Society of New Zealand.

Music: Masterclasses delivered using the Access Grid

In mid-2008, New Zealand Symphony Orchestra cellist Brigid O'Meeghan will be delivering four masterclasses to University of Otago students using the Access Grid. Because the Access Grid uses the high bandwidth provided by the KAREN network, participants can collaborate musically in real-time.

Lico de Ridder, AV Technical Specialist in ITS Teaching Services, is working with participants to find the best setup. "I used a quality cardioid mic, our 3CCD digital video camera, and angled lighting at this end. The setup needs perfecting but the participants were very positive about the trial."

cello_lesson_1.jpg
 
VUW logo
Contact Us | Section Map | Disclaimer | RSS feed RSS FeedBack to top ^

Valid XHTML and CSS | Built on Foswiki

Page Updated: 23 Jul 2008 by sam. © Victoria University of Wellington, New Zealand, unless otherwise stated. Header image used and relicensed under Creative Commons. Original author: whurley.