Info on F@H work units - NOT A 72 HOUR TIME OUT!

[bunker]

Here are the protein descriptions.

The majority of them have a 6 or 7 day time out, not a 72 hour time out like most of the older proteins.

 

[Stanman]

I think Stanord benchmarks each system to determine suitable work to give it depending on its speed.
I have also heard that if you are over the limit, you still get credit for the work, they just don't use it, so it's just good for the stats and not the cause.

 

[nagger]

I've been checking some of the proteins that I've folded and none was within the 72h limit.

Some of them can last up to 6 days so the 72h limit doesn't apply here.

Probably has something to do with being the F@H 2.0...

 

[Poof]

bunker - the thing to remember is that these proteins are different sizes, thus just because it has a longer "time out" period, doesn't mean that a slow processor will make it in time. In general, about the minimum processor speed for the project would be a P2-300 or K6-2/333.

And with respect to result rejection, unless they changed it recently (which would be good), people were not getting credit for WUs turned in beyond the time frames set.. The ones who noticed it were necessarily the stats freaks...

In all their communications, the F@H folks recommended G@H for slower machines.

Maybe someone might be willing to try returning one past the limit to see if it is credited or not.

 

[randomlinh]

nm