The Folding@Home project, which has recently been thrust into the limelight by the gaming press, was created by professor Vijay Pande at Stanford University. A few of us at TPS decided that we should find out what's going on over at the Pande group, and who better to ask than the guy in charge? Mr. Pande was pressed for time, so please excuse the brevity. Richard Marks of SCEA's research and development team was kind enough to make a guest appearance, and he clarified for us just how Folding@Home stresses the hardware of the system.

TPS: How has the Folding@Home project evolved from its early days?

Pande: We are looking at dramatically more challenging systems, moving from basic biophysics, to direct simulation of fundamental biology and disease. Our work on Alzheimer's Disease is a great example of this.

TPS: What is your opinion of other distributed computing projects? Do you feel that they spread the available CPU power too thin between organizations, or do they simply offer different approaches to similar problems?

Pande: Distributed computing is a natural and powerful way to do research. When we started Folding@Home in 2000, there were no other projects studying biology or chemistry or physics (there were projects like SETI and distributed.net). Now, there are numerous science oriented projects. Our early success was a demonstration of the power and usefulness of distributed computing in these new areas and so it's natural that other researchers would want to use these methods.

TPS: Although Folding@Home has produced some wonderful data and presented new information to researchers, some people have expressed a desire for more tangible results. What do you have to say to them?

Pande: There are many tangible results (eg see our results at http://folding.stanford.edu/papers.html). If you mean drugs, one should keep in mind that even if we happened to find an interested inhibitor of Abeta aggregation (the cause of Alzheimer's Disease), it would take 10-15 years to bring that through the whole drug design process. On the short term, our productivity in papers and our international recognition (see http://folding.stanford.edu/awards.html) is a strong sign that we're making significant progress.

TPS: As the F@H project gains more users and increased exposure, do you plan to compute more complex simulations that were not possible before?

Pande: Absolutely. As we get more computer power, we can do more complex and challenging simulations.

TPS: Who made initial contact between the Pande Group and SCEA to bring the F@H client to the PS3?

Pande: There was mutual interest, but they found me first.

TPS: Is the PS3 more suited towards certain kinds of protein simulations?
If so, in what situations is it not an ideal use of resources?

Pande: The Cell is well suited to the calculations we're running on it and can get some extreme speed increases due to it. Nevertheless, the PC CPU's are a bit more general, allowing for a greater diversity of calculations.

TPS: Are you concerned that interest in the project from PS3 users will decline after the first few weeks?

Pande: We expect some amount of decline. That's the nature of how this works.

TPS: Do you plan to update the F@H client for the PS3 in the forseeable future,
or is it currently everything you had hoped it would be?

Pande: Absolutely. We expect there will be regular updates.

TPS: Are you looking to expand F@H's use of other technologies and platforms, such as with your recent GPU and SMP efforts?

Pande: Yes. We took what we learned from the GPU client to design the calculations to run on the PS3.

TPS: Lastly, do you play any games in your free time? If so, what are some of your favorites?

Pande: What is this "free time" you speak of? :)
Life's been pretty busy, but in the past, I've been known to play Starcraft and other similar RTS games.



We thought it would be best to save the more PS3 specific questions for an SCEA employee, and that's where Richard Marks comes in. Here's what he had to say about the safety of running F@H for long periods of time.

TPS: Is it safe to run the PS3 24/7 for folding, whenever it is not being
used for gaming/movies/etc?

Marks: Yes, it is safe to run the PS3 24/7.
We have not received any indication of problems with the PS3 running for long periods of time.

(Marks noted that the PS3 should have adequate ventilation so as to not overheat.)

TPS: If so, should users still give the PS3 a break to cool off, and how
often?

Marks: There have been no indications that the PS3 needs any cooling period when
running the Folding@home application. PS3 has a built-in fan that regulates its internal temperature.

TPS: What internal components are stressed the most heavily during folding,
and how might this affect the stability of the system?

Marks: The main component of PS3 that is stressed during folding is the Cell
processor. The Cell was designed as a high-performance processor for just such purposes.
There is no expectation that Folding@home will have any adverse affect on PS3 stability.

TPS: Did the Folding@Home project pull any resources away from other projects
like, say, the HD EyeToy that has been mentioned at the past couple E3s?

Marks: No, not at all. :)

I assume you ask this because of my role in helping to promote Folding@home. It is actually being developed by another team here within R&D, but I have helped spread the word because it is such an important initiative, and I have some previous experience promoting new concepts. My personal area of focus continues to be new interface technology,
but I am honored to be an advocate for the Folding@home project.



We'll be sure to let you know of any F@H updates in the future.