Folding@Home Now More Powerful Than World's Top 7 Supercomputers, Combined
The fight against coronavirus goes nuclear
Propelled by average enthusiasts in their shared quest to defeat COVID-19, the Folding@Home network is now pushing out 470 PetaFLOPS of raw compute power. To put that in perspective, that's twice as fast as Summit, the world's fastest supercomputer, making the network faster than any known supercomputer. It's also faster than the top seven supercomputers in the world, combined.
It's impressive that the Folding@Home network is now more than twice as powerful as Summit's 149 PetaFLOPS of sustained output: ORNL announced two weeks ago that Summit had also joined the coronavirus fight and has already found 77 different small-molecule drug compounds that might be useful to fight the virus. Summit employs 220,800 CPU cores, 188,416,000 CUDA cores, 9.2PB of memory, and 250PB of mixed NVRAM/storage for the task.
But Summit is far faster than the other supercomputers further down the Top500 list. That means the Folding@Home network is also now faster than the world's top seven supercomputers, combined. That's equivalent to the horsepower of 27,433,824 CPU/GPU cores that are being used in the most powerful systems in the world. These leading supercomputers are typically only used by nation-states for decidedly more nefarious purposes, such as nuclear research, so seeing this type of compute power unleashed for the common goal of defeating the coronavirus is certainly encouraging.
As promised, here is our first glimpse of the #COVID19 spike protein (aka the demogorgon) in action, courtesy of @foldingathome . More to come! pic.twitter.com/iD2crCMHcXMarch 16, 2020
Here is a view of the enemy, stunning in its complexity, and deadly in its intentions. This virus may be sweeping the globe, pushing large portions of the world into isolation at both the national and personal level, but the global community is coming together through the Folding@Home network to fight back by furthering research into possible cures or vaccines. This consists of using your computer to complete small chunks of much larger problems, thus giving researchers access to an unprecedented amount of compute horsepower.
Distributed computing has always been a great hobby because of the detailed stats compilation and the dizzying number of teams involved, but Folding@Home's addition of coronavirus research to its normal pursuits, like cancer, Alzheimer's, and Parkinson's research, has led to an overwhelming amount of new users. Folding@Home reports that it has seen a 1,200% increase in contributors, with Bitcoin miners also joining the fight, and over 400,000 new volunteers have joined over the last two weeks.
Unfortunately, that massive surge in demand has led to a shortage of work units (the small chunks of larger workloads sent to each user), but Folding@Home has expanded its capacity to serve units to speed production. Work units are still being issued and many more are in the pipeline.
You can help, too, by simply installing the Folding@Home application and turning over some of your spare CPU or GPU horsepower to help defeat the virus. It only takes a few minutes to set up the program, and then it's effortless as the program runs in the background.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
We're contributing, too. Tom's Hardware and our arch-enemy/sister-site AnandTech are currently embroiled in a month-long contest for folding supremacy, which you can join by following the simple steps outlined here.
Since @PaulyAlcorn asked, contributing an 8x Tesla P100 system (more here https://t.co/hbiXRj5CQ3 @NVIDIADC ) to @foldingathome #COVID19 efforts. Also a fun note, I saw that I am still a top 6000 contributor after having been a top 20 years ago in the quad @AMD Opteron G34 era pic.twitter.com/OrFzAwzYbeMarch 21, 2020
ServeTheHome recent joined the Tom's Hardware team with an impressively-fast Gigabyte G481-S80 server armed with a whopping eight Nvidia Tesla P100-SXM2's running for the cause.
But you don't need that type of horsepower to make a meaningful contribution – every single bit helps. It's entirely possible that some old-school Pentium somewhere will calculate the bits that serve as the final piece to solving this deadly puzzle. We're issuing a call to arms to our readers: Any additional help to battle the coronavirus, not to mention AnandTech, would be appreciated. You can join the team here.
Paul Alcorn is the Managing Editor: News and Emerging Tech for Tom's Hardware US. He also writes news and reviews on CPUs, storage, and enterprise hardware.
-
penn919 I just enlisted my testbench setup which I installed spare parts onto. I hope my Athlon II x2 coupled with an UBER powerful GTX 460 1GB can sway the tide...well, maybe not.Reply
When I first tried, the GPU would not queue any work. Eventually, I got an error message complaining about not finding opencl.dll. I found a publicly available file and placed it in the client's appdata folder. That seems to have solved the issue. -
PaulAlcorn penn919 said:I just enlisted my testbench setup which I installed spare parts onto. I hope my Athlon II x2 coupled with an UBER powerful GTX 460 1GB can sway the tide...well, maybe not.
When I first tried, the GPU would not queue any work. Eventually, I got an error message complaining about not finding opencl.dll. I found a publicly available file and placed it in the client's appdata folder. That seems to have solved the issue.
Thanks for the effort! Don't worry, every cycle counts! -
SkyBill40 I belong to a close knit gaming community and we created a team in order to participate as well. Any efforts are certainly better than none.Reply -
bit_user It really needs GPUs, the faster the better.Reply
I'll see if I can get my Radeon VII working, in a few days. This will be a good chance to "break it in".
First, I have to do something about the fan curves, as I'm now sitting in the same room with it all day. -
nammi-namm penn919 said:I just enlisted my testbench setup which I installed spare parts onto. I hope my Athlon II x2 coupled with an UBER powerful GTX 460 1GB can sway the tide...well, maybe not.
When I first tried, the GPU would not queue any work. Eventually, I got an error message complaining about not finding opencl.dll. I found a publicly available file and placed it in the client's appdata folder. That seems to have solved the issue.bit_user said:It really needs GPUs, the faster the better.
I'll see if I can get my Radeon VII working, in a few days. This will be a good chance to "break it in".
First, I have to do something about the fan curves, as I'm now sitting in the same room with it all day.
According to https://apps.foldingathome.org/project?p=14328 Coronavirus is a CPU only project. So beefy GPU's are really only helping with non-Coronavirus related projects. -
derekullo nammi-namm said:According to https://apps.foldingathome.org/project?p=14328 Coronavirus is a CPU only project. So beefy GPU's are really only helping with non-Coronavirus related projects.
"These projects are CPU projects to simulate the main protease of the COVID-19, a possible drug target. These supplement high-priority GPU projects of the main protease as well as the COVID-19 receptor binding domains."
Both the CPU and the GPU are valuable for Folding@Home COVID-19 research.
The CPU test is focused only on simulating the main protease of COVID-19.
The GPU test is focused on simulating the main protease of COVID-19 as well as understanding which receptor binding domains to target.
A protease is an enzyme that catalyzes (increases the rate of) the breakdown of proteins into smaller polypeptides or single amino acids.
https://en.wikipedia.org/wiki/Protease
Receptor binding domains are the areas that the virus binds to on a cell to begin infecting that cell.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3754068/
The above link is for the another epidemic back in 2012 but it was also a coronavirus, MERS-CoV.
https://en.wikipedia.org/wiki/Middle_East_respiratory_syndrome-related_coronavirus -
Zizo007
Radeon VII is a beast in computing, I am sure its faster than 2080Ti in compute.bit_user said:It really needs GPUs, the faster the better.
I'll see if I can get my Radeon VII working, in a few days. This will be a good chance to "break it in".
First, I have to do something about the fan curves, as I'm now sitting in the same room with it all day. -
egearbox This is seriously amazing. I'm glad to see that humans really are capable of working together.Reply -
Tigerhawk30 I've only just now heard of F@H, and only through this article...it appears it's been around for quite a while. How did I miss this for so long...???Reply
I'm kind of fuzzy on what's needed, what kind of resources are being used in the background, etc since I can't seem to find anything that boils it all down and rather seems more interested in the hardware vs the processing sides. Even the site itself is extremely sparse for information on what resources the program actually uses, just that it runs in the background, and otherwise the info I've found only seems to be geared towards people who've been doing it for a side "profession", as it were.
Essentially, my very basic dummy question is (and I hope I'm not sounding "greedy/selfish" here)...if I'm gaming using my 2700X/RX 580 8GB combo, does allowing F@H to run in the background affect anything? Or, maybe in the interest of the reverse, does gaming affect the processing of F@H? How much in the way of resources would I be taking away from F@H processes when I use it for my own purposes? What about when Win10 wants to install updates then reboot to apply them?
I'd have no problem letting my system run 24/7 if it'd help. Just curious what kind of slowdown it would cause on both ends when running games, both single player and MMO.
Thanks in advance!