Did you know: the first machine to break the exaflop barrier (one quintillion floating‑point operations per second) wasn’t a huge dedicated IBM supercomputer, but a bunch of interconnected PCs with ordinary CPUs and gaming GPUs.
With that in mind, welcome to the [email protected] project, which is targeting its enormous power at COVID-19 research. It’s effectively the world’s fastest supercomputer, and your PC can be a part of it.
Put simply, [email protected] runs hugely complicated simulations of protein molecules for medical research. They would usually take hundreds of years for a typical computer to process. However, by breaking them up into smaller work units, and farming them out to thousands of independent machines on the Internet, it’s possible to run simulations that would be impossible to run experimentally.
Back in 2004, Custom PC magazine started its own [email protected] team. The team is currently sitting at number 12 on the world leaderboard and we’re still going strong. If you have a PC, you can join us (or indeed any [email protected] team) and put your spare clock cycles towards COVID-19 research.
Getting your machine folding is simple. First, download the client. Your username can be whatever you like, and you’ll need to put in team number 35947 to fold for the Custom PC & bit-tech team. If you want your PC to work on COVID-19 research, select ‘COVID-19’ in the ‘I support research finding’ pulldown menu.
You’ll get the most points per Watt from GPU folding, but your CPU can also perform valuable research that can’t be done on your GPU. ‘There are actually some things we can do on CPUs that we can’t do on GPUs,’ said Professor Greg Bowman, Director of [email protected], speaking to Custom PC in the latest issue.
‘With the current pandemic in mind, one of the things we’re doing is what are called “free energy calculations”. We’re simulating proteins with small molecules that we think might be useful starting points for developing therapeutics, for example.’
Bear in mind that enabling folding on your machine will increase power consumption. For reference, we set up folding on a Ryzen 7 2700X rig with a GeForce GTX 1070 Ti. The machine consumes around 70W when idle. That figure increases to 214W when folding on the CPU and around 320W when folding on the GPU as well. If you fold a lot, you’ll see an increase in your electricity bill, so keep an eye on it.
Folding on Arm?
Could we also see [email protected] running on Arm machines, such as Raspberry Pi? ‘Oh I would love to have [email protected] running on Arm,’ says Bowman. ‘I mean they’re used in Raspberry Pis and lots of phones, so I think this would be a great future direction. We’re actually in contact with some folks to explore getting [email protected] running on Arm in the near future.’
In the meantime, you can still recruit your Raspberry Pi for the cause by participating in [email protected], a similar project also working to help the fight against COVID-19. For more information, visit the [email protected] website.