Discussion in 'Folding@Home' started by BUCK NASTY, Feb 1, 2011.
Thanks for joining/helping Mydog
just set my brother XFX 260 black edtion to fold for chimpowerup, we needed to stability test before selling so wat better way then fold for TPU aye kiex gave me the idea
Wow, we jumped to just 2 points behind OCN, looks like we got a shot at 2nd place! Keep on pumping everyone!
Does the PPD look right for bigadv on my rigs?
Someone here suggested that I was just running normal SMP's but I'm pretty sure I'm running -bigadv WU's.
I'm switching the rest of the rigs when they finish their WU's
Looks good... P6900,P6901 work units are -bigadv running the A5 core
EDIT: which app are you using to monitor your rigs, I like the look of it, is that HFM
I know the WU's are -bigadv but are the PPD correct with the bonus?
HFM is correct, works perfect for me.
The 970,980 compare well with mine...I get ~60K PPD running at 4.3
Cant help with the SR-2, I just wish I had one to compare with
The others dont look too far away in PPD as well
No, not so much as how to OC the card to get more out of it, but if I needed to check -advmethods or anything like that in the client.
Looking fine Mydog. Second place here we come
1.54 chimp points behind ocn now we should be able to make a strong push now to pass them forsecond with everyones help
Time to push the OC on my chip, I think I can squeeze out another 500MHz.
My 980X has dropped down from 64K PPD to 51K. Is that normal, or is it a sign of instability and needs to reboot?
Probably a different WU with a lower point value.
Anyone knows why specific projects keep on failing on my 9600GSO cards on my folder #1 (specs below)? 5767 (IIRC) for example always dies a second after it starts with "UNSTABLE_MACHINE", while 577X projects work just fine. Is there any way to force it to do work for specific projects (or to avoid others) ? The machine is stable as a rock CPU-wise (5 hours prime95 and 20 rounds of LinX) and the GPUs are at stock. The 8400GS doesn't seem to have this problem, and I can't babysit my folders all the time.
The most common culprit is a bad WU:shadedshu Believe it or not Standford puts them out there not knowing how there are going to perform Graphics cards are different so your 9600GSO might not like it but my GTS250 will do fine with it. It's just the way that WU is coded IIRC. Not sure what client you are using but with the GPU2/GPU3 console clients, you can re-configure this by changing the WU size that your client receives.
Might it be a lack of GPU memory issue? The 9600GSOs are 384MB cards, while the 8400GS which has no problem is a 512MB card. Maybe it simply runs out of memory on the GSOs ?
Running GSO's at stock is almost criminal.
I have an almost identical setup to yours and I don't have issues with those WU's.
The memory is not an issue. The 384MB GSO's were the backbone of many farms, including BUCK's. I also run a 256MB 8800GT without issue.
The project numbers that you gave represent 353 point WU's. Today, these and the 587 point WU's obtain the highest PPD. I've also seen that the 353's run hotter and pull more electricity than most of the other WU's available. (I get 4420 PPD with my GSO's)
I have to agree with Bog. Sometimes a GPU doesn't like a WU.
Do you have a molex plugged into the "additional power connector" on the motherboard?
Yep. I'll clock the cards later when I have the time to do some stability testing. Life is a bit hectic right now.
I once had a -bigadv that went from 21:30 to more then a hour per frame for 3 frames. Nothing else was running so, yes it can happen that PPD goes up and down a bit during the same WU. Just leave it as is.
Edit: 1.66 M the first 12 hours of today. Seen worse
Chimp Challenge 2011 End of Day 6 Stats
The dance between OCNChimpin and ChimPowerUp continue but with our newly discovered firepower I have no doubt that ChimPowerUp will secure the #2 position.
In the last 24 hours the active clients for ChimPowerUp increased from 297 to 333.
My DSL was down for 6hrs last night, lost some production, but all is fine now.
We are on pace today (EOC) to hit 3.27 million points for the day! Already over 2 million and 3 updates to go.
Funny seeing my PURE teammate and me folding for different teams in the chimp challenge
This may sound odd, but I'm happy to report that my computer crashed* .
I know that having a computer crash isn't usually good, but I desperately needed to clean the fans and I didn't want to lose my four month 100% uptime record. I called in the people who dug out the Chilean miners to unearth the heatsinks buried under the dust piles on my motherboard.
Now I can fold on four cores 24/7 instead of just 2 during the day and 4 at night. Additionally I can update my prehistoric CCC 10.8 drivers to 11.4p and maybe start using GPU3 with my HD 5830.
I managed to gather a few more volunteers from amongst my friends. I should have them set up and folding some time this evening.
*Crash was due to a RAM error apparently. Fortunately it's lifetime warrantied.
Separate names with a comma.