TPU's F@H Team

[mstenholm]

Got a big one finally

 

[roast]

Give it a try. I run 2 9600GSO's without dongles, but on XP which I thought was worse concerning the need for dongles. You could also try the V7 client. https://fah-web.stanford.edu/projects/FAHClient/wiki/BetaRelease You will not see a PPD increase as the HD5xxx and newer will see. Also, at this time, I do not know of a third party monitor that works with V7.

Cheers mate. tbh, a monitoring client is a must for me. I'm rarely at home, so I like to see how things are doing. I'll just hold out and pick up a dongle at the weekend.

 

[johnspack]

I've tried on 2 intel and 1 amd mobo with 2 cards, under both xp64 and win7 64, and have never needed a dongle! Think I've even done it in sli mode....

 

[roast]

I've tried on 2 intel and 1 amd mobo with 2 cards, under both xp64 and win7 64, and have never needed a dongle! Think I've even done it in sli mode....

Never had such luck unfortunately. Tried in SLI mode, failed, tried without dongle, failed.

Just tried it on the new clients there, same thing. Not to worry.

 

[TheoneandonlyMrK]

id like to ask an ernest question, ive the rig in my sig basically 5850 and 5870 no cpu client on that one, soon to be fully wc and both running at a gig whilst folding plus a pentium D at 3.2 with a gt240 both folding all running v7 and optimised accordingly, this is the question bit

is this equipment best used here or crunching,
im more interested in work done then points tho im not against points its just i saw yet another loved one buried today and work done seems more concerning to me ie which would i get the most work done for Fahome or crunching thanks

obv team TPU either way

 

[johnspack]

Brought my old amd rig back to life, so sputtering out some ppd on 2 260s now. Have my 2nd i7 cpu now, to go with 2nd x58 mobo. So now to bring to life my first ever i7 folding box, I need a 12gb kit of preferably 1866 ram, so I can give the folding rig my 6gb kit, and I need a good cooler, preferably a nice used cpu water cooling kit. I'm paid up on the cpu now, so I need offers for the last 2 things.....

 

[thebluebumblebee]

id like to ask an ernest question, ive the rig in my sig basically 5850 and 5870 no cpu client on that one, soon to be fully wc and both running at a gig whilst folding plus a pentium D at 3.2 with a gt240 both folding all running v7 and optimised accordingly, this is the question bit

is this equipment best used here or crunching,
im more interested in work done then points tho im not against points its just i saw yet another loved one buried today and work done seems more concerning to me ie which would i get the most work done for Fahome or crunching thanks

obv team TPU either way

IMHO, crunch on the pentium D and Fold on everything else. And OC that GT240.

 

[Athlonite]

-bigadv takes a minimum of 7 cores

with ver 6.xx maybe but as far as I know F@H ver 7 now enforces the 8 core rule for bigadv

 

[Athlonite]

Got a big one finally

I'm sure that's what you think she said but in reality it was Bung not big

 

[johnspack]

Well, gotta keep up my sad sack folder, going to crawl, but will get that 16th spot! Can't give up when I'm so close.... And again, need offers for a 1366 water cooling kit in particular, and also a 12gb kit of 1600 or better ram. Can do a first payment through paypal on the 23rd, and may need to do a 2nd a month later. That's why I ask in here first.... It will result in a full time i7 folder. Any offers at all....?

 

[TheoneandonlyMrK]

IMHO, crunch on the pentium D and Fold on everything else. And OC that GT240.

ty i will turn them to that purpose i oc everything to the max stable folding speed but at mo its turned down a bit due to midnight noise, i take it you think low core cnt cpus get better results crunching( for future ref)

i stopped folding on the quad because its low ppd impacted the gfx cards and wasnt worth it

 

[thebluebumblebee]

ty i will turn them to that purpose i oc everything to the max stable folding speed but at mo its turned down a bit due to midnight noise, i take it you think low core cnt cpus get better results crunching( for future ref)

Short answer-yes. Midnight noise? Some on this team can't sleep without the sound of whirling fans.

i stopped folding on the quad because its low ppd impacted the gfx cards and wasnt worth it

The AMD GPU client still hits the CPU hard. So your low PPD on the Q6600 was because of the GPU clients. I seem to recall that an OC'd q6600 could do 10-12k PPD with Nvidia GPU(s) Folding. Therefore, combine the q6600 with the GT240 and the P4D with the ATI cards, idle the P4, done. Max Folding PPD. j/k

 

[johnspack]

I agree, intel quads can do 10k ppd easy. My q9450 with 2 gtx260s was a powerhouse, no impact on ppd on anything. Not sure what ati cards do to it. I saw up to 15k ppd on some days just with the quad.

 

[Jstn7477]

One ATi client uses 12% of my i7-2600K so having a few ATi cards going can really throw a wrench in your CPU PPD. NVIDIA is much better with not only better PPD, but less CPU usage at the same time.

 

[theonedub]

Picked up a entry level GT530 to add to my system. Thought it was a good low power card to put in the PCIe x4 slot in my computer- for about 45w it will produce around 3500ppd @ stock 700c/1400s.

 

[bogmali]

how do you join?I don't know how to?

What hardware do you have? Go to the main folding page and check out the folding essentials tutorial to get started.

 

[xBruce88x]

Welll I've decided to add my Baked 88000GT to the folding mix. Figured it was about time I got back to folding lol. Since this is the only folder i have (aside from a 8600gt... that needs baked) I figured it should at least perform a little better than the 9600. I've got it overclocked to 650core, 930mem, and 1620shader. could probably go higher but i'm not gonna risk it with stock cooling... though I have the A/C on to help with that lol. It's still running a test work unit (just got it set up) so i'll post back when I get an idea of the PPD. I just used the GPU3 client from stanfords site, the one with the fancy viewer, and I also loaded up the latest version of FahMon. I don't use the viewer much, but its nice to see from time to time. Its using the GPUv2 core though (since my card isn't a Fermi)

It's been a while lol...
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&teamnum=50711&username=xBruce88x

 

[thebluebumblebee]

^^ OC'ing a G92 in order of performance for Folding: Shaders, Core, Memory. I run my 9600gso's at 600/1725/don't care, which I think is very conservative, but there is no manual fan control and I like to keep the temp under 70C. My 8800GT does not have the cooling of the GSO's so I can't run the clocks as high. You will get the most PPD with the GPU2 client on the pre-Fermi cards.

 

[xBruce88x]

Mine's an OEM card. I'm just using the nvidia control panel for overclocking and forced the fan to max (noisy little bugger) and turned on my A/C unit lol. Right now with the gpu maxed out on load, its running at 58.3C

If you really want to, you can take the power wire on your fan and shove it into the yellow port of a 4pin molex on your PSU, that'll force it to max speed lol

 

[KieX]

Looks like some initial numbers are available for the new 3960X in F@H.
43K PPD on SMP @ 240Watts (Stock settings)

Source

 

[Chicken Patty]

That's pretty sick, right?

 

[KieX]

That's pretty sick, right?

Perhaps those with more experience with the Westemere/Gulftowns can answer that. I'm not too familiar with SMP/Hex

 

[TheoneandonlyMrK]

just a thought but a while ago i had a gt240 in my main rig hybrided up for physx in batman, but had to remove it, because it kept nicking the x16 folding job off the 5870 every now and again, but its now folding in a seperate rig and so im pondering, since its open cl compatible and did the job at one point( i watched it fold) can i force it to run x16 work units now as it did them pretty quick for a gt240

 

[mstenholm]

Perhaps those with more experience with the Westemere/Gulftowns can answer that. I'm not too familiar with SMP/Hex

As posted in the pie section - I think it is on a 609x WU. TPF was 5:23. My 970 @ 4.15 does it at 5:55 so there is some improvement.

 

[KieX]

As posted in the pie section - I think it is on a 609x WU. TPF was 5:23. My 970 @ 4.15 does it at 5:55 so there is some improvement.

Oops, sorry missed that