Catch Giant Space Explosions with Gaming Technology

Two science institutions from the Netherlands – NWO Institute and Netherlands Institute for Radio Astronomy (ASTRON) - were in need for more computing power to measure huge explosions in space. So they build their own supercomputer using components from the gaming industry.

The Crab Pulsar and the Crab Nebula

The Crab Pulsar and the Crab Nebula. This image shows a 1-second snapshot of peaked waves as they reach ASTRON's Westerbork telescopes after 6000 years of interstellar travel.

Image: ASTRON/ESO/AURA

Outer space can be violent, unleashing thousands of giant explosions daily. Space scientists call these Fast Radio Bursts (FRBs) which can generate enormous energies and flashes. These bright radio flashes travel billions of light years to reach Earth. Their origin is still a mystery due to the fact that they only last a fraction of a second, which makes them difficult to spot. Over the past 10 years only 25 bursts have been detected.

Supercomputer

In order to measure these forces of nature, NWO and ASTRON installed new high-speed cameras on radio telescopes in the Dutch town of Westerbork. The device is equipped with one of the latest models of sensitive wide-field cameras, Apertif. The camera continuously shoots high-speed movies of the radio-sky at 20,000 frames per second. Processing this many images requires a little more power than your average home computer. "To form and process all those images, we need the computing power that only the fastest supercomputers in the world can produce", says Joeri van Leeuwen from ASTRON and the University of Amsterdam. "But we did not have such a computer yet. That's why we designed and built this one ourselves."

Their new supercomputer is named ARTS. Where regular supercomputers use super high-tech components – ARTS uses off-the-shelf image processing chips from your average gaming computer to do its dirty work. As van Leeuwen explains: "Gamers use very powerful processors for video tasks: the GPUs. We now use these chips for the first time to process the high-speed images of our telescope." The Dutch supercomputer supports 200 of these GPU-chips.

Self-learning

The Dutch computer will have the ability to teach itself to spot FRBs out of thousands of pictures. "We have always done this manually. But that is a lot of work, and also error-prone. As it detects more flashes, the supercomputer will learn to distinguish the flashes better and better. We hope to discover one FRB per week. Our system is unique in the world in that it can then also immediately determine the precise location," says van Leeuwen.

Using the ARTS supercomputer will bring the world’s scientists closer to understanding Fast Radio Bursts.