Page 1 of 1
Slow progress
Posted: Sat Nov 10, 2012 5:23 am
by Leda1974
My friend and I have a team (just the two of us so far) and I've noticed a massive disparity in our work rates. She has managed to complete 30 work units in the same time it's taken me to complete just 2. Both of us are each only running F@H on a single laptop; it's probable that her machine is more powerful than mine (hers is a Macbook and I just have a basic machine running Win7), but can it possibly be that much more so? Is there any way I can improve my work rate?
Thanks for any advice.
Re: Slow progress
Posted: Sat Nov 10, 2012 5:58 am
by P5-133XL
I cannot see how your question can be answered without far more information. At the very minimum we would need to see the client logs and know the HW involved.
There are large scale differences in HW capability as well as WU's vary greatly in the amount of work they do. It is possible to have a 15:1 ratio in WU's finished and everything be properly optimized. Can't tell without details.
Re: Slow progress
Posted: Sat Nov 10, 2012 6:03 am
by 7im
Yes, it can possibly be that more powerful. It can possibly be 5 other reasons as well. As P5 said, without more detail, we can't even make good guesses.
Also note she might not use the laptop that much, so it folds more and quicker, while you may be a real study bug, so fah has to share the laptop resources with you.
And note that not all work units are the same size, or the same points. 30 small workunits can equal 3 large work units. But again, we're only guessing without more detail.
Re: Slow progress
Posted: Sun Nov 11, 2012 4:00 am
by bruce
Even if the two computers turn out to be the same speed (and we certainly don't know that yet), that doesn't mean that all WUs are equivalent to each other. If the WUs are from different projects, they can be working on proteins that are radically different in complexity. Take a look at the
Project Summary showing the active projects. Some have maybe 500 atoms and others have more that a million. That's why FAH invented the concept of points. Projects are benchmarked on points are assigned based on the amount of work in a single WU of that protein.
Bonus points are also also awarded based on how quickly you complete the WU, but that's another subject entirely.