Folding@Home Was Just Getting Started - ExaFLOP Barrier Breached as Network Achieves 1.5 ExaFLOP Performance
It's not been that long since we praised the worldwide community for allowing Folding@Home to reach double the performance of the world's top supercomputer, IBM's Summit, which ekes out calculations at the petascale with 200 peak petaFLOPS. However, the world has again shown promise when it comes to people selflessly giving their power and performance for a higher cause. Folding@Home now likely smokes performance metrics of some Tier 1 civilizations out there (in the Kardashev scale, of course, should there be some; I'm cautiously jesting in all regards) with its 1.5 ExaFLOP performance.
As it stands, and should users keep on donating their processing power, it seems the community can even go as far as eclipsing the deployment of the world's most powerful supercomputer, El Capitan, which is only meant to be operational - get this - in 2023. This means we are close to offering equivalent performance for multiple disease research (including COVID-19), today, to that of a currently theoretical supercomputer. Of course, queues are long towards receiving work, so now it's become a problem for researchers to keep this computational power fed with data to crunch - we're actually hitting bottlenecks that supercomputers too have achieved before us.
As it stands, and should users keep on donating their processing power, it seems the community can even go as far as eclipsing the deployment of the world's most powerful supercomputer, El Capitan, which is only meant to be operational - get this - in 2023. This means we are close to offering equivalent performance for multiple disease research (including COVID-19), today, to that of a currently theoretical supercomputer. Of course, queues are long towards receiving work, so now it's become a problem for researchers to keep this computational power fed with data to crunch - we're actually hitting bottlenecks that supercomputers too have achieved before us.
