Imagine instead of sitting through the same boring old stress test software runs over and over again to test your new hardware, you could sit back and let your PC process radio telescope observations of dying stars. That’s exactly what Pawsey Supercomputing Research Centre in Australia has been doing with its newest machine, Setonix, in order to reveal one of the most breathtaking images of a supernova’s remnants I’ve seen in all my years of stargazing.
Named after possibly one of the most adorable animals in the known universe, the quokka (opens in new tab) (Setonix brachyurus), the AMD-powered Pawsey supercomputer is anything but adorable. The team put this new machine through one hell of a benchmark using data gathered by the 36 dish antennas of CSIRO’s Askap (opens in new tab) (Australian Square Kilometre Array Pathfinder) radio telescope, and boy did it prove to be robust (via The Conversation (opens in new tab)).
Setonix was tasked with processing highly complex data—sent through a series of high-speed optical fibres—into separate images spanning hundreds of different frequencies. It then had to combine the images into their final form.
With all the data, it managed to pull together an image of the imaginatively named SNR (supernova) G261.9+05.5, which was originally brought to CSIRO’s attention way back in 1967. The supernova is somewhere in the region of 13,000 light-years from our cosy home planet.
What I find really sweet is that the Setonix machine itself was decorated by artist Margaret Whitehurst of the Wajarri Yamatji people, and was appropriately “inspired by the stars that shine over Wajarri country in Western Australia’s Mid-West,” the site (opens in new tab) reads.
But enough about art, back to science.
Not only was the data selected especially to test the supercomputer’s hardware (which it passed with flying colours) it was also able to determine the clout of the Askap processing software, ASKAPsoft. If you’re curious, Setonix is an EX model supercomputer from HPE Cray (opens in new tab), and consists of the following:
- More than 500 AMD EPYC “Milan” CPU nodes (65k cores total) – 64 cores, 2.55GHz, 2 per node, 256GB per node
- Eight 1TB High Mem CPU nodes
- Eight Data mover nodes
- Sixteen Visualization nodes
- Four Login nodes
- Connected by HPE’s Slingshot interconnect (100Gb/sec)
- Lustre file systems (14 PB [3 SSD, 11 HDD])
Sadly the rest of us are unlikely to match that kind of hardware, there are ways you can use your PC to contribute to science. Software such as [email protected] (opens in new tab) may not be able to stress test your PC, but it can help the scientific community to find cures for cancer, ALS, Parkinson’s, Huntington’s, influenza, and others.
You can also track asteroids and potentially help study the origins of our solar system with [email protected] (opens in new tab), if you’re looking to get involved with something a bit more out-of-this-world. Why not give it a go?