TPU's F@H Team

100k on a 5770??? I got one of those sitting in a drawer.

I do have the beta flag on. Im not sure what they are.

 

They are normal core 17 WUs. I get 87k PPD on a OC 7950 (950 Mhz) and 123k on a OC 7970 (1070 MHz) so these numbers are unbelievable good

 

hmmm down to 59k now i guess my celebration was short lived lol

Now 15 minutes later back up to 102k. anybody else notice the huge ppd swings on 13000?

 

I thought those were gone. I got a P9401 (100,0,23) on my 7770 which bumps it down to 21,602 PPD
and @Norton beat me to the P13000 wild swings.

 

I got a 86 yesterday (9401) so yes they are still around. My 7950 is now and every time I checked the past few hours doing 115k and my 7970 is doing 105k, both on later runs so there are still some difference between the WUs, not like the 8900s. You could set your clock from the time of completion from these guys. My 280X has been stable at 149k (1111,0,41) the first 56%. That will make up for the hours of waiting for a 13000/8900 for two rigs until I set the beta flag. Sorry beta testers - I took them from you.

 

The 13000 WU I'm running on my R7 250 seems to be jumping between +-20% of the actual TPF. Not the first time I've seen FaH WU's do that.

 

13000 and the 9401 both do that the first few runs. The three 9401s that I run now are +/- a few % from the final TPF but the first two-three runs were all over the place. On the 14th of March I posted my early experience with the 13000s - TPF fluctuating a lot, FAHControl later learned to calculate it right.

 

Four 8018 rejected by server. Wished they would get this fixed.

 

They have tons of problems it seems. No beta units, no 13000s and the 8900s are off the shelf and then your 8018 problem which I haven't come across on the F@H forum. I got three GPUs doing nothing but sucking idle power. Yes I know that they know about the lack of WUs but it seems to take forever to load new ones.

 

I think F@H needs some down time. I'm calling for a 1 week boycott starting April 1st. What do you think fellas? Do you think we could get a few teams, especially the big teams to go along with this?

 

I doubt it but they sure deserve it. Did you read the blog? Plenty of cooperate GPUs just started folding and they keep upping the TERAFLOPS but either these cooperate GPUs get all the work or the TERAFLOPS are BS. Anyway I'm as pissed as you are (read your F@H post , totally agree with you).

 

folders Folding@home

 

My 770 was pulling 110k on a 9401 and that last 680 I got is golden was getting 130k at 1293 core on a 9401. So all these guns and nothing to shoot at

 

I noticed that you got some decent points today. Did you run your GTX 680 at 1293 MHz (boost)? Not bad. My 280X pulled 143k PPD on a 9401 yesterday . Today it is closer to 0 but at 300 MHz.

 

You should have read the part that I didn't post!
What if we tried for a one day boycott? April 1st. April Fools Day.

 

Yea but not on purpose computer rebooted and precision started them both at same settings and since they are different cards one was at 1223 and the new one was at 1293. It had been that way for a least 4 hours so I figured what the heck Ill leave it there. The 770 Im still playing with 1230 right now. Not bad for a dead card.

 

I'm in but with the speed PG work we might have an involuntary 14 day "boycott".

Edit: I have informed the Danish Electricity Board that they can take out their big power boiler for maintenance since the expected load will go down for a longer period.

Edit: got to much time on my hand - http://folding.extremeoverclocking.com/user_summary.php?s=&u=610327. I think he is testing his new toy...

Edit: BBB you got another one supporting your disgruntle on F&H. I long time ago learned that today means at best in the afternoon, after lunch the same but more likely tomorrow, shortly means "I have no idea how long it takes", shortly the same. I noticed that that Avedøre 2 (mayor power boiler in DK) is now down

 

Well, so much for the 480 I tried to fold, thought it might be okay so I put it back in my main rig to sli again, it went into endless reboot, so I pulled it quick, went back to normal. Put my server back together with my old 9800, but it was doing same thing. Found out the bios is messed now, ddr3 options go from 3000mhz to 9000mhz, so it only boots on auto now. At least I pulled it out of my main rig before damage occurred there. Got my new 27" monitor today too, so bittersweet. Can't believe one video card caused so much damage.....

 

Well now is the time to test things. We just got the Indians interpretation of soon which is later today. In my world that is tomorrow, maybe.

Edit: Yes, 4-8 hours is his new prediction. I had to deleted my reply to that or I would have been banned for live and pulled to court for inflammatory remarks. Over and out for today

 

Well, I have to try to reflash my server, then look for proper folding cards. 480s really aren't a good idea.....

 

Beta is up and running, Run 36 and 103 for me. Somebody must have read my remarks on the efficiency of PG

Still almost 5000 to go.....

 

You lost me.????

 

http://fah-web.stanford.edu/pybeta/serverstat.html Here you have all the servers and the amount of Wu ready to send out. Atm 4985

Edit: server IP 171.67.108.31 row #32 and now 4860, get them while they are hot..

Edit: 13000 is server IP 140.163.4.231 and there is open for business as well with close to 5000 WUs.

 

GTX 660 or better. I personally feel that 660 Ti's are the best $/PPD. On the AMD side, 7770 (just because they're twice as efficient as GTX 560 Ti's and do the same PPD) or better. Virtually all GPU work is now QRB based so better to get one "better" card than 2 "lesser" cards.
Refer to this chart for further guidence: https://docs.google.com/spreadsheet/pub?key=0AmQ6YZGMTUpFdHhxNkptVncyR2k4dVV6WDBmQXRJYVE&gid=0