Seriously, where did all the GPU Work Units Go?
Folding@Home, a distributed computing project started by Stanford University, is one place where people can donate computer time to help scientists cure disease. The project recently announced that new models were being developed to help researchers understand the latest coronavirus that is wreaking havoc across the globe.
As a result, tons of new people have downloaded the F@H client and started chewing away at molecular dynamics problems. This is great, because it means a lot more geeky computational charity is happening. Just check out the graph below. X-axis is time. Y-axis is performance. When people realized they could join one of the world’s largest supercomputers to fight COVID-19, BOOM! Instant quadrupling of computer horsepower. That’s sweet.
But, as it turns out, there’s a twist. All those new computers ate up all of the existing work units, so now the F@H Consortium’s servers have run dry.
That’s right, as of the writing of this post, it is impossible to get GPU work units, just like it is impossible to get toilet paper, hand sanitizer, and anything made by Clorox. For people who incidentally use Folding@Home to heat their house in the winter, this is really annoying!
Here’s a screen shot of one of my seven empty Folding@Home computers…just like those shelves at the supermarket, there’s nothing here.
The Folding@Home forums are rife with people noting this problem as of 3/14/2020.
Here’s the official announcement:
Thankfully, just as my local supermarket recently announced a new shipment of toilet paper, F@H has announced that more COVID-19 projects should be hitting the streets. So, here’s to hoping these come my way soon, so that I can fight this virus with my computer as well as with my hand sanitizer (yes, I have some, but I’m not telling you where).
Update: 3/14/2020 at 9:00 PM
I’ve got heat in my bedroom again from my small 1070 Ti based space heater. So, I went downstairs and found that my dual GPU benchmark machine is up and running with fresh work on both the 1080 Ti and the 980 Ti. Here’s to hoping this makes a difference, and that the scientists behind the project can benefit from all this added computational capacity (and keep the WUs flowing!).
Thank you to all the donors (veterans and new COVID-19 donors alike) and the F@H researchers and volunteers!