The momentum and excitement for the Folding@home project continues. This time it’s something that the Folding community and the computer science field as a whole have been anxiously awaiting — the crossing of a milestone known as a petaflop.
A petaflop equals one quadrillion floating point operations per second (FLOPS). If you’d like to imagine this enormous computation capacity, think about calculating a tip on a restaurant bill, now do that for 75,000 different bills, now do that every second, and lastly, imagine everybody on the planet is doing those calculations at the same second- this is a petaflop calculation. Now you see why I say say enormous…
So how did we get to a petaflop and what caused it to happen now? There are a few factors that together increased the total computational power of the Folding@home network. First, is the increased participation from the Folding community. Just six months after we launched the program, nearly 600,000 PS3 users have registered. Second, we made several improvements to the application (v 1.2) that helped make the computations more accurate and enabled us to squeeze even more work out of each and every PS3 console — we went from 450 teraflops to 800 teraflops. These factors, combined with the contribution from all the other platforms, helped us cross the barrier, which happened sometime over the weekend.
What does this mean to the computer science field? The petaflop by itself is just a number, but in the past was only thought of as something that could be done on a supercomputer or other dedicated hardware which cost millions and millions of dollars. Now we’ve shown that a distributed network featuring a videogame console can reach the same capacity as a supercomputer. On a different level we see a huge contribution to the specific research that Folding@home is conducting, which is understanding the cause of serious diseases. Just think about how long it would take if such a network of computers didn’t exist! According to Stanford, the Folding@home network is helping them conduct research that typically would not have been possible for another 10 years.
What’s next? 2 petaflops? Maybe. Researchers at Stanford are saying that the accuracy can be increased even further to crank even more work out of each PS3. For now, I just feel proud that PS3 owners are making such a great contribution to the network and allowing it to sustain such capacity, it makes all our efforts worthwhile.
Here is a snapshot of statistics as reported from Stanford’s website at the time of this post: