Supercomputer boosts brain research
Case study: Cubric is directing the action
Researchers at Cardiff University are using their supercomputer to significantly improve the efficiency of their brain analysis work.
The Cardiff University Brain Research Imaging Centre (Cubric) claims to have the largest high-performance computing cluster in the UK.
The team has been using the supercomputer for 10 months to support its neuro-imaging and data analysis research.
Previously, images had to be processed consecutively but the new computer can process 100 complete brain images simultaneously in just 16 minutes - 24 times faster than before.
Want more photos?
Click here to browse the full archive of our photo stories.
Spiro Stathakis, IT systems manager for Cubric, told silicon.com: "The lead time for getting a study done has decreased dramatically."
He explained that this means that the research can now include a greater number of subjects, which in turn means more accurate statistics can be produced.
Stathakis added: "We've got a greater possibility of getting meaningful data."
The supercomputer also allows the team to overlay MRI images of the brain with MEG (magnetoencephalography) images which show the electromagnetic fields indicating neural activity.
Stathakis said: "It [the imaging] needs quite of lot of processing."
Due to the supercomputer's greater speed, it has also become easier to fix problems when they occur as not as much time is lost.
Cubric has been running the supercomputer for 10 months but only now has it become been clear how much it has benefited the team.
Stathakis said: "We had to make sure it was bedding in smoothly, we had to get people in through the door. We haven't had any significant problems."
The supercomputer runs on Linux meaning the team has been able to modify software code for use with the system, which has also taken some time to complete.
Although it is used primarily for neuro-imaging, other departments in the university – such as Chemistry and Earth Sciences – have also made use of the machine.
As the MRI tech is able to scan the whole body, the engineering department has been able to use it in the development of a new prosthetic limb.