This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Distributed computing with MATLAB speeds development of International

20 March 2006

The Distributed Computing Toolbox from The MathWorks has been used to speed the development of the International Linear Collider (ILC), a proposed new electron-positron collider that would allow physicists to explore energy regions beyond the reach of today's particle accelerators. The toolbox enabled multiple computational intensive simulations of the accelerator and particle beam to be carried out in parallel for the first time with a high level of realism, effectively reducing development time a hundredfold.

The collider was modelled and simulated with MATLAB and Simulink, which helped to produce accurate simulation results,
and then executed in a grid cluster using the Distributed Computing Toolbox. The ILC design incorporates two 20 km-long linear accelerators that hurl beams of ten billion electrons and positrons toward each other at nearly the speed of light. To accurately design the operational control systems required to bring the beams, which are just a few nm-thick, into collision, a simulation was designed to precisely model  the way the beams travel through the accelerator.

This simulation needed to be repeated multiple times to model random imperfections and arrive at an optimal design, a demanding task that took a standard PC with a 3 GHz CPU approximately two-to-three days for just one ‘seed’ or configuration.

Using the Distributed Computing Toolbox, the simulations could be run in parallel on a grid computing cluster, typically enabling 100 seeds to be processed in the time it would normally take just one.

“The Distributed Computing Toolbox significantly reduced the computational time of each simulation, enabling us to carry out
numerous different tests that would not have been possible otherwise,” said Glen White, Physicist at Oxford University. “We
used MathWorks tools because they are simple to use, the learning curve is always very fast when using a new toolbox, and the documentation for MATLAB products is excellent.”

The ILC team’s current objective is to develop a reference design report for the accelerator by the end of 2006. White now wants to extend the use of the Distributed Computing Toolbox over the grid of 12 universities in the UK involved in the project. Depending on international funding decisions, the ILC could potentially be
operational by 2016.



Contact Details and Archive...

Print this page | E-mail this page