Folding@Home: THGC Needs You -Team 40051

Page 425 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.


The devils canyon chips did go back to the old solder on heat spreader. One of the members on overclockers tried to delid his and destroyed the die in the process.

Look like I missed a really good show with this Forcha44 fellow.
I always keep popcorn and a Pepsi ready for these events.
Oh well maybe next one. :pfff:

 


i never oc my gpu for folding and its the ist x21 iv had trouble with.NEW POWER SUPPLY !!!:no:
 
^Nope

I have had multiple failed units over the past week on both the 980 and the 980ti.

It is purely on the wu's that start 96**, never anything else.

I have had a steady stream of 0x21 wu's and only these ones have issues.

Both cards are at stock settings, I just cant wait fro these wu's to run out!!!

00:35:53:WARNING:WU00:FS02:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
00:35:53:WU00:FS02:Sending unit results: id:00 state:SEND error:FAULTY project:9631 run:0 clone:8 gen:62 core:0x21

08:11:28:WARNING:WU00:FS02:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
08:11:28:WU00:FS02:Sending unit results: id:00 state:SEND error:FAULTY project:9630 run:0 clone:40 gen:17 core:0x21

06:58:14:WARNING:WU01:FS02:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
06:58:14:WU01:FS02:Sending unit results: id:01 state:SEND error:FAULTY project:9639

23:45:20:WARNING:WU00:FS02:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
23:45:20:WU00:FS02:Sending unit results: id:00 state:SEND error:FAULTY project:9626 run:1 clone:23 gen:61 core:0x21

Feel my pain Pimp🙁
 


:lol:
 


thought the new psu sorted it.im doing anouther finishing in 3hr 48 mins so we shall see
 
It won't help. That is why I said "or that you have a GTX 960, 970, 980, or 980Ti". There is something about Maxwell V2 that makes those units fail. Throw it on a Kepler and you won't see a single failure. Hopefully it will be fixable in SW or drivers, but we'll see. I wish they would restrict those projects, they never should have been released to full fah imo as there are far too many maxwell v2s out in the wild.

Note: Sometimes they will finish successfully. Normally they won't.
 


It really sucks, especially considering how well the 900 series cards fold otherwise. I'd love to take 96XX WUs all day if it would let people with 980s etc fold on all the other projects that work fine. I get good PPD and zero errors from them.
 
My biggest frustration with the stalled WUs is that it can take a few hours sometimes for FAH control to catch it and revert to the last checkpoint. This results in a quite a bit of lost production.
 
thought the new psu sorted it.im doing anouther finishing in 3hr 48 mins so we shall see
Click to expand...

The new PSU sorted the lack of supply to the cards, unfortunately from what EXT has said and further investigation this problem is on-going, pity because the 0x21 units don't half churn up the ppd.

At least with the failures the returned units still get some points.

 
Yeah, reducing memory clocks did help on some other projects and does seem to influence the rate of BS, but it wasn't enough for 96xx. Hopefully they will restrict that one while the next phase of testing to figure out what is going on occurs.

The issue was up until recently a lot of their development testing and benchmarking occurred only on Kepler, where these run beautifully, so the seriousness of the Maxwell bug was not recognized as quickly (even though it was caught in internal testing).
 
lol i been thinking about another 8 pack 980gtx. this thing [now iv learnt how it all works is an animal and runs cool baby. it looks as if SidVicious hold on the top 30 spot is weakening to the power of me[as the national grid can testify]] wish id waited for black friday and grabbed a 980ti hof version. still. you can just keep spending and spending with these things.
 
you can just keep spending and spending with these things.
Click to expand...

Yep...

Now check this out, folded 0x21 units all night fine and you get 0x18 errors!!

00:48:50:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
00:48:50:WU02:FS01:Sending unit results: id:02 state:SEND error:FAULTY project:9135 run:235 clone:0 gen:15 core:0x18 unit

07:51:00:WARNING:WU01:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
07:51:00:WU01:FS01:Sending unit results: id:01 state:SEND error:FAULTY project:9136 run:360 clone:0 gen:3 core:0x18 unit
 


Don't sweat it mate, I am beginning to think it has something to do with the AS at Stanford. I have been getting bad Core 17 WU's recently and the last time that happened there was mention on the F@H forums as well of servers being taken offline and the way that the servers allocated work going all wonky. There was also mention of how bad WU's were re-farmed out to see whether the problem was with the WU or the machine trying to process it.
 
Yep, a few random bad WU is nothing to sweat. And come on pimpanther, you'll make NVidia so happy at their next earnings report if you just keep spending, spending, SPENDING! 😀 😀 😀

Seriously though, I don't plan to buy any more GPUs until the next series is out. That should be a big jump in performance and efficiency after sitting on the same process node for so long.
 


You and me both brother! :)
 
but everyone was waiting for the new amd and ehat a flop that seems to have been,8gb of memory but still performs normally against nvidia card.i nearly bought a 390 but this 8 pack thing saved the day phew glad i saw it.it looks so sexy with its black shroud and hall of fame in bright white light..and a world of OC to do with it.ceramic compasitors etc,some saying they have hit gpu clock at 1500mhz and its possible with an OV to get it to 2000.[i wont be doing that lol
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom