I know how Stanford work out the scores for each folded work unit, based on their benchmark p4 2.8 - but was wondering....

If a PC had to be logged off for say 12 hours a day, thereby doubling the actual amount of time, do you score half the points? I assume FAH just see a date/time for a WU going out, and a date/time it gets sent back and works out the points on that?

Soooo, the main question is, "Do you get better scores from leaving a PC on 24/7?" (than doing it twice as long if machine needs shutting down)

Anyone know?

Not sure Jason, I would have to guess it's real processing time, not actual over time.

Oh update sig as the /cms stuff will be disappearing soon

I'm just smacking my head as my laptop was not runing the high risk wu's...grrr

Some of the WUs have expiration dates.. if its not finished in time you wont get any credit..

the points per WU is fixed, no matter how long it takes to do it.. now my opteron is crunching thru roughly 2 a day

the points per WU is fixed' date=' [/quote']

aha. I wondered as it was always based on time, if you got it in late it wouldn't count. Thanks for that.

Yeah. What Colin G said.

What I can't understand is how they benchmark. They can't run every single work unit on their benchmark machine - otherwise there would be no point in farming them out to distributed processors. So. do they benchmark once per protein "family", or is it related to the particular processing core (tinker, gromacs, etc.)?

It's not important - I'm just inquisitive; it would be nice to know