Stanford uses a P4 2.8 with HT off to now to benchmark wu's. In the past it was a 500 mhz celeron for benching. The reason why there are folks out there with far more wu's and less points is the structure of how the points are calculated has changed sever times. When I started folding in march of '02, there were no gromacs wu's, only tinkers. The largest tinker unit was worth 10 points and would take 20 hours on a xp @1800 mhz to fold (bout 60 percent longer on a P4 of equal clock) A very common project at the time though, was a tinker that took 6 hours on same rig and was worth .6 points. That 1.8 ghz rig on average would bang out about 70-80 ppw which at the time was kickin arse. Then gromacs were introduced, which kinda horked everything up for the points system. The gromacs core allows for far greater efficiency in the amount of work that can be accomplished per clock cycle, and could take advantage of SSE instructions. This resulted in stanford at first making gromacs worth far more in points per hour than tinkers. Then stanford came out with huge tinkers, and increased the pph there although still not equal to gromacs. and lately stanford raised the tinkers to nearly equal gromacs in terms of points per hour. in the interest of trying to not **** off anyone, any time there was a change in the points system, the pph would always be moved up. so wu's that take a set time are worth far more than say 2 years ago. It kinda screwed the folks who folded early on who folded many wu's but got few points. But at least now, the work is fairly even as far as pph go and is getting closer every time they make a change. I hope this helps explain it bit although i left quite a few things out along the way.
