folding@home wu's failed

[tallun123]

hi guys ive just been checking on my folding progress and ive noticed that gpu1 has been failing to do its wu's 6806 the log says [unstable machine] i cant understand this as smp and gpu 0 have been ok ive had no bsod or crashes of any sort checked the log for smp n gpu 0 every wu has completed i only have the cpu overclocked both gpu's are @ stock

it failed 10 wu's in a row gutted thats 13500k ive stopped all work n restarted it an so far all is fine is there any reason for why gpu 1 is failing ? as always any help is much appreciated thanks guys

 

[tinytomlogan]

if its overclocked it may well be unstable dude, thats the first thing to look at

 

[stevehat1]

Be sure the gpu flags are correct, -gpu 0 for card 0 and -gpu 1 card 1. also be sure they are machine 2 for card 0 and machine 3 for card 1

 

[tallun123]

its only gpu 1 thats failed smp n gpu 0 are fine on the same system its running again now it says in the log [unstable-machine] failed 10 wu's paused for 4 hrs if the overclock was unstable would it not fail on all the folding cores? like i say smp is fine been running in exsess of 36hrs + along with both cards its running again now @ 35% fingers crossed it will finish possably a glitch on stanfords side ? this setup has been rock solid stable @ 4.2ghz now for weeks prime 95 linx it flies though 3dmark vantage ive had no issues at all with crashes bsod just seems odd how its all of a sudden started failing on this one card [gtx 460] only today just these 10 fails in a row

like ive said in a previous post my temps are aok 60c full load im going give it more time to see if it sorts its self out n1 tom for your quick reply

 

[tallun123]

Be sure the gpu flags are correct, -gpu 0 for card 0 and -gpu 1 card 1. also be sure they are machine 2 for card 0 and machine 3 for card 1

hi m8 im useing the fah tracker so it configs everything for you doesnt it ? how do i check which machine is running which card? thanks

 

[tallun123]

doin my head in now gets to 50% then fails n starts again i cant understand it gpu 0 [gtx 570]is running 6806 not a problem and gpu 1 [gtx 460] is also running 6806 gets to 50% and fails ive got avdmethods enabled for the gpu3 client could this be my problem?

 

[kup]

Disable AdvMethods and let Tracker autodetect the cards and Clients again. It might even be a driver issue, what revision are you running?

 

[tallun123]

im useing tracker 3.40 everythings been sound untill today i'll try disableing the advmethods n restarting the tracker do i need to delete the old wu's from the directory location?

 

[kup]

Yeah, you could even delete the entire DIR and start again if it's not playing nice. You only need the Tracker .exe to start again.

 

[AMD_PBz]

what everyone said my general consensus is the new WU's stress the cores more and i've had to back down my overclocks to keep both 460's stable just remember prime and vantage stable does not equal Folding@home Stable.

 

[stevehat1]

hi m8 im useing the fah tracker so it configs everything for you doesnt it ? how do i check which machine is running which card? thanks

Keep forgetting that everyone uses the tracker, mine are all individual client setups.

You might try deleting the work folder if it happens again. Also switching the cards around might to be something to try....

 

[tallun123]

n1 guys ive just started off a fresh the machine id's are 2,3 n 4 dont know where 1 is ? point taken about stability but this problem has only happened today . strange

anyways ive checked all the logs n things seem to be back on track hopefully happy folding dudes

 

[tallun123]

maybe this is something tom can do one of his video's on how to configure folding@home clients properly ive tried a couple of times now but without any success a video tutourial would be mint

 

[AMD_PBz]

this may help if you've not seen yet how to setup Tracker

 

[tallun123]

i hadnt seen it thanks its the way ive done it its still doing it aswell after a new install

im going to take the graphics cards out and re seat them n do a fresh install again to see if that works i suppose its back to default clocks n try it to see if it is the overclock i think its more to do with me swapping 1 of the 460's for my new 570 even though it ran all night hmmm started messing up from 2pm today i installed the 570 about 4pm yesterday ?

never mind

 

[tallun123]

just been thinking it wont be anything to do with the evga precision tool will it as the 570 evga n the 460 is gainward its just that the tool controls both card n was wondering weather it may be causing the fault on the gainward 460??

 

[kup]

It could be. You can edit the settings to control each card individually.