Author

Topic: Why does one of my cards drop out after a while? (Read 710 times)

member
Activity: 266
Merit: 10
Hey guys thanks heaps for the replies.

I'm using powered risers on all 3 cards.

And I have set the clocks back to default on the unstable card.


I've been meaning to get the Pro Btc H81 board but they were out of stock. I'm gonna preorder one in tomorrow and hopefully have it by next week.


Could it be really that this mother board can't take 3 x 280x's ?? :/
zvs
legendary
Activity: 1680
Merit: 1000
https://web.archive.org/web/*/nogleg.com
I guess its your motherboard too..

yeah, probably right, unless he's messing with the core/memory settings of the card as well

a powered riser should fix that problem
newbie
Activity: 51
Merit: 0
Hello Good people Smiley


I have just gotten a third Sapphire Dual-x r9 280x

and added it to my set up.  ( Asrock H81M-Plus )



I've managed to get them all running at ~ 715, 719, and 750 khash/s
which I guess is ok..


But the third card seems to drop out after about 45 mins or an hour after starting cgminer.

The temp drops and khash/s drops out to 1.5 , 1.5 mhash/s


I've not iced I sometimes get a message saying "GPU2 idle declaring sick GPU attempting to restart. Idle for more than 60 seconds"

I've undervolted the cards (using Trixx) 2 are on 1.168  , and now I'm trying to set the voltage back to default on the problem card but it still seems to be dropping (it just dropped to 300khash/s)

I'm running a 1000W Silverstone Gold PSU, and have 4GB of ram in the system.



Does anyone have any clues or ideas what it could be? Would gladly appreciate any help Smiley

thanks  Cool

or your motherboard might not provide enough power for three cards

I guess its your motherboard too..
zvs
legendary
Activity: 1680
Merit: 1000
https://web.archive.org/web/*/nogleg.com
there are a bunch of new BIOSes @ http://www.asrock.com/support/download.asp?cat=BIOS   didn't see one specifically for your model, but H81M-HDS and H81 Pro BTC ?
zvs
legendary
Activity: 1680
Merit: 1000
https://web.archive.org/web/*/nogleg.com
Hello Good people Smiley


I have just gotten a third Sapphire Dual-x r9 280x

and added it to my set up.  ( Asrock H81M-Plus )



I've managed to get them all running at ~ 715, 719, and 750 khash/s
which I guess is ok..


But the third card seems to drop out after about 45 mins or an hour after starting cgminer.

The temp drops and khash/s drops out to 1.5 , 1.5 mhash/s


I've not iced I sometimes get a message saying "GPU2 idle declaring sick GPU attempting to restart. Idle for more than 60 seconds"

I've undervolted the cards (using Trixx) 2 are on 1.168  , and now I'm trying to set the voltage back to default on the problem card but it still seems to be dropping (it just dropped to 300khash/s)

I'm running a 1000W Silverstone Gold PSU, and have 4GB of ram in the system.



Does anyone have any clues or ideas what it could be? Would gladly appreciate any help Smiley

thanks  Cool

Uh, if a card "drops out" and doesnt totally freeze or crash your system, that would indicate to me that it's semi-close to being stable, but the settings need a bit more changing.

I would say to check the VRM temps, but that doesnt really make sense if it has more issues (3khash) when you're undervolting it than when you leave it at default (300khash).

so, i guess you are probably changing the core or memory settings?  one of them is moved over too far

or your motherboard might not provide enough power for three cards
member
Activity: 266
Merit: 10
Hello Good people Smiley


I have just gotten a third Sapphire Dual-x r9 280x

and added it to my set up.  ( Asrock H81M-Plus )



I've managed to get them all running at ~ 715, 719, and 750 khash/s
which I guess is ok..


But the third card seems to drop out after about 45 mins or an hour after starting cgminer.

The temp drops and khash/s drops out to 1.5 , 1.5 mhash/s


I've not iced I sometimes get a message saying "GPU2 idle declaring sick GPU attempting to restart. Idle for more than 60 seconds"

I've undervolted the cards (using Trixx) 2 are on 1.168  , and now I'm trying to set the voltage back to default on the problem card but it still seems to be dropping (it just dropped to 300khash/s)

I'm running a 1000W Silverstone Gold PSU, and have 4GB of ram in the system.



Does anyone have any clues or ideas what it could be? Would gladly appreciate any help Smiley

thanks  Cool
Jump to: