• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

New WU's Killing PPD...

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

prime81

Member
Joined
Feb 8, 2008
Location
Odessa, Texas
Anybody else experiencing this trend? I was averaging about 4000-6000 PPD and now it's drop significantly due to these new WU's. Instead of 2653, all of my machines are grabbing 3060 WU's.
 
Running a single instance on a Q6600, most of the 30xx SMP WUs return nearly the same ppd as p2653. Those running on a dual core or running 2 instances on a quad, will see a pretty dramatic drop in ppd. These WUs may prove to be a good reason to spend the extra money on a Q9450 or better. Cache contention is likely the culprit for the ppd decline.
 
yai got some massive WU lately in the 15000000000000000000000000000 range but i am only runing them on dual cores, taking almost an hour to completle one section
 
yai got some massive WU lately in the 15000000000000000000000000000 range but i am only runing them on dual cores, taking almost an hour to completle one section

Yeah I too got a couple of those massive WU's even on a Dual they take anything from 30 -60min /step and only got something like 50pts
 
I have one box that's getting nothing but them pesky things! C2D's don't like them.
 
I'm just running the single core client on my Pentium Ms and I've been getting a lot of nasty little WUs. They take a fair amount of time to complete but I end up with 4 points for each of them!!!
Makes me miss the usual Amber core 1500 or 5000 frame WUs.
...oddly enough these little ones were nowhere to be found during the March Madness Folding Contest :beer:
 
Computer on right side of sig (dedicated folder) previously been doing SMP units at around unit/day/1760 points. Got a 2605 three days ago, checked this afternoon, hadn't got the first frame. Figured a bad unit, so removed the queue and work dir, restarted, it downloaded a new 2605 unit, and 9 hours later, no first frame. Bad batch of units from Stanford or what? My Phenom system that I use for gaming (also running SMP client) is getting 3064 units for a bit now, and doing the same unit/day/1760 points. What's up with the 2605 units on Linux SMP?

Just checked my EOC stats, looks like the first bad unit got to my machine 4/17 or 4/18...

EDIT: hank123, if your reply next post was directed at this post, I'm not complaining about points. I'm complaining about seemingly broken SMP units. The second 2605 units on my C2D machine is now at 12 hours and not a single frame. Something is wrong with these units. Two in a row shouldn't be broken like this on a machine that was previously eating a unit per day.
 
Last edited:
BRING THEM ON !!!!

T32 Folds ANYTHING !

The sooner we get through them the sooner we get something better !
We've been here & done this B4 guys.


FoldOn!
:attn:
 
I'm just running the single core client on my Pentium Ms and I've been getting a lot of nasty little WUs. They take a fair amount of time to complete but I end up with 4 points for each of them!!!
Makes me miss the usual Amber core 1500 or 5000 frame WUs.
...oddly enough these little ones were nowhere to be found during the March Madness Folding Contest :beer:

Sorry my boxes have been stealing all those little guys. I'm running 2 clients on two seperate computers till I take the time to figure out running 2 on my dual core then I'll have 3 but so far in the last 3 or 4 days with a combined processing power of 3.80ghz I've gone through 21WU's all night tonight I've been getting either 150 frame or 500frame's. I don't know how to get the bigger ones otherwise I'd take a stab at them.

Aha, just after I typed this I finally gave SMP a try and the first one my dual core grabbed is a 3064 but I've got just shy of two full days until the preferred deadline so I might make it.
 
Last edited:
Computer on right side of sig (dedicated folder) previously been doing SMP units at around unit/day/1760 points. Got a 2605 three days ago, checked this afternoon, hadn't got the first frame. Figured a bad unit, so removed the queue and work dir, restarted, it downloaded a new 2605 unit, and 9 hours later, no first frame. Bad batch of units from Stanford or what? My Phenom system that I use for gaming (also running SMP client) is getting 3064 units for a bit now, and doing the same unit/day/1760 points. What's up with the 2605 units on Linux SMP?

Just checked my EOC stats, looks like the first bad unit got to my machine 4/17 or 4/18...

EDIT: hank123, if your reply next post was directed at this post, I'm not complaining about points. I'm complaining about seemingly broken SMP units. The second 2605 units on my C2D machine is now at 12 hours and not a single frame. Something is wrong with these units. Two in a row shouldn't be broken like this on a machine that was previously eating a unit per day.

If you check the log, you'll find you probably downloaded the same WU after deleting the Work dir and queue.dat. The AS and Work Servers are programed to send a machine the same WU in the event of a "lost" WU. When your rig went back for another WU, without returning the last assigned, it sent you the same one. You pretty much have to fold the bad units twice to know they're bad (they'll die in the same place) and then delete Work and queue until you get a different WU. I'm not seeing many bad p2605s, BTW, and as for ppd they are about identical to p2653. And why let it go 12 Hours without progress? 43:12/frame and you won't make the prefered deadline, 57:36/frame and you won't make the final.
 
Last edited:
Back