[b]LASPHEMY!!!

Hey... I'm just joking :D Those brackets are all that is evil :p

Just finished sifting through the F@H official forums and found out a a couple of things. I am not 100% certain of any of these and haven't tested them myself so do not rely on this as some of the users in those forums seemed dumb, not all of them, just a few.

1) F@H Folds faster in Windows than it does in Linux as the client was optimized for Windows.

2) Regarding folding on a P4 w/HT: I gathered that running a single client seems to fold faster than two clients. (Per 1 CPU)

If anybody else finds (if they have time ;) ) any relevent information regarding F@H help us all out and post it :D

It is about 15% faster to run 2 seperate instances with HT enabled with most of the protiens. You just have to make sure they are listed as seperate machine IDs when you do setup.

One person on the F@H forums said that running two was slower with HT enabled... have you tested this? Goes to show you can't believe everybody, sorry for the misinformation :(

I know it's luck of the draw on which protein you fold.

Why do I seem to be getting the hardest/longest proteins???
At about 132 points per WU nobody else is even close.

I'm not really complaining, just wondering.

Yes I have tested it. I have 5 HT enabled P4s folding. The worst I have found was a couple that had no gain but most gain about 15%.

Here is what I have folding right now.

http://webpages.charter.net/chew_toy/1/EMIII.gif

I'm getting strangely easy proteins that are worth jack. To put this in perspective, my 8 proteins got me 222 points, and Murray's 9 are well over 1000 (more like 1200). Incoherent's 3 proteins got him 228 to my 222. I need to get my friends to fold but they're worthless. They won't even run their systems 24/7 because they 'can't sleep with the noise'. I tell you, they whine like little girls.

Allright,,, I'm signed up and folding!:)

Ive been taken over... :eek: :eek: :eek:

WTG Chew_Toy :D

What method did you use to benchmark your machines?

I tested with actual protiens I was folding and times per frame. That is the only way to know, some protiens that have the same batch number and points fold at different rates.

I have 4 machines folding now.
P4 2.4
P4 2.53
P4 2.6
P4 2.8 o/c 3.57
need to find more machines to fold :evilaugh:

Were almost in the top 1000 :D w00t

Not to shabby for having just started folding... not that we are going to make it to the top though :( Damn [H] and their huge concoction of idiot folders :mad:

:cool: We're 999!!!

Does anybody know which yields better results? WU's less than 5 megs vs WU's greater than? I switched a few machines over to smaller WU's and I am now getting on average 57 points per WU. When I had them set to WU's larger than 5 megs I was getting on average 35-40 points per WU. Is this just a coincidence?

*edit* We're now in the 800's :evilaugh:

For those who like graphs
It looks like we need to have a team ranking of 613 or better to start getting individual stats. That should happen within an hour or so. :D
More stats (if you're not looking for per-user graphs)

On the HT front; my (limited) testing seems to show nearly 100% improvement by running 2x clients per HT CPU on Linux 2.4.20-8 (i.e.: no slowdowns on per-frame times vs. running single client). Will update when I parse more logs later.

For EOC it is top 500 that have the individual scores. Statsman has them for the top 1000

http://www.statsman.info/folding2stats/39339.html

The EOC stats will show more details and is closer to actual time for the updates than Statsman's site. We should be in the top 500 next week.

Wow, there must be a lot of teams moving up ranks rapidly. I had just done some casual clicking and noticed that teams ~rank 600 seemed to have indiv. stats.

We're 553!! Should be 500 next Tuesday, possibly sooner.

Okay, I'll help get us there faster and sign up with 2 home machines and 4 work machines (All P4 3.0+) :D

Top 500 keep it going!

Those stats are cool... We should be in top 300 in a month or so...

It will take a while for the individual stats to give somewhat accurate info tho. A lot will be ok a week from sunday after the weekly stats reset this sunday and get a week of true stats.

or mine - which have all 380,000+ individual stats for the past 28 days...