Folding Forum

Hi,

I'm relatively new here, so I apologize if this is the wrong forum or if there is another thread on this topic. I realize Folding@Home is a bit different from other DC projects. My question is how many WUs does it take to reach a quorum (or are there no quorums here)? I look at my PS3 stats for example and I see that I have folded this particular WU 32 times. I've been looking at the Project Summary page for a few months, and it looks like the GPU clients are still only folding this lambda repressor, and the PS3 clients are still folding the supervillin.

Is there a number of times a WU has been folded when it is deemed "complete" (e.g. 100? 300?) Will the GPU clients ever fold something other than lambda repressors, or the PS3 clients fold something other than supervillin?

Thanks

The concept of quorum doesn't apply to FAH. As a general rule, all WUs need to be folded ONCE and there's no expectation of duplication, (although it does happen, at times, in order to insure completion of "lost" WUs that later turn up). There just aren't enough computers participating to waste resources on duplication.

A unique WorkUnit has four values, Project, Run, Clone, and Gen. (Unfortunately for those with a PS3, the software doesn't display the numbers for Run, Clone, and Gen -- only the project number and name -- but the GPU does.)

To understand concepts of Run, Clone, and Gen please read the explanation here in the Wiki. If you still have questions after reading that, come back and ask them.

In general terms, a Folding@Home Work Unit does not represent an entire 0-180 degree fold; your machine might calculate a few degrees worth, then someone else picks up where you left off. Repeat until all 180 degrees have been finished. Breaking each project into these small chunks keeps the overall progress moving along, and makes it easier to detect when work has been lost.

There's a much more detailed explanation here, if you're interested in the finer points:
http://fahwiki.net/index.php/Runs%2C_Clones_and_Gens

edit:
Beaten!

Just wondering how much more folding will be spent on the "villin headpiece subdomain"? It seems that is all the ATI and PS3 (and nVidia, but understand that one more since nVidia is new) clients have been crunching, for quite some time now (many many petaflops). And yes, I've read the explanation of runs/clones/gens. Thank you.

esoteric17 wrote:Just wondering how much more folding will be spent on the "villin headpiece subdomain"? It seems that is all the ATI and PS3 (and nVidia, but understand that one more since nVidia is new) clients have been crunching, for quite some time now (many many petaflops). And yes, I've read the explanation of runs/clones/gens. Thank you.

You may or may not have read Dan's explanation here: viewtopic.php?f=17&t=78&p=32757#p32757

As I see it, the critical words are

. . . we hope that with enough calculations, even with all the complexities and approximations, that we can come up with *general principles* of protein folding so we don't have to calculate them all. I'm not likely to figure this out myself this afternoon, let alone in the next two years . . .