970 crashing issues

[Nasher]

I'm having strange problems with a 970 where sometimes I get a black screen, usually followed by a lockup.

After it occurs there are 100s of Event ID 13 logged in the same second all with different variations of graphics exceptions. The first one has "Injected Bundle Error", second has "Class 0xffff Subchannel 0x7 Mismatch". etc. all seemingly different.

It appears to happen in some games more than others. But I have not been able to trigger it with Furmark.

I've tried it with a different MB, PSU, ram, various driver versions, even monitor and it still happens. Anyone come across it before?

 

[harney]

I'm having strange problems with a 970 where sometimes I get a black screen, usually followed by a lockup.

After it occurs there are 100s of Event ID 13 logged in the same second all with different variations of graphics exceptions. The first one has "Injected Bundle Error", second has "Class 0xffff Subchannel 0x7 Mismatch". etc. all seemingly different.

It appears to happen in some games more than others. But I have not been able to trigger it with Furmark.

I've tried it with a different MB, PSU, ram, various driver versions, even monitor and it still happens. Anyone come across it before?

Had same thing similar with a 660 & 970 you could try and re flash gpu with the stock manufactures bios

It worked for me

 

[Nasher]

As in a non-OC BIOS? Not sure there is one for this card. It's a Gigabyte 970 OC mini, they only made one model afaik :/

 

[harney]

As in a non-OC BIOS? Not sure there is one for this card. It's a Gigabyte 970 OC mini, they only made one model afaik :/

No just reflash the bios with the same but not the one from your card if you know what i mean....

had a few cards with black screening and also some where i could not install the display driver correctly
card would not be found only stranded vga but once i re flashed the gpu bios all was good


its worth a try at least

 

[Holst1981]

I'm assuming temperatures are ok?

Have you run memtest and cpu benchmarks to rule oou cpu or ram.

Although the error looks like a gpu problem it's worth ruling out the cpu and ram.

 

[harney]

What bios does gpuz say you have

 

[Nasher]

Temps are ok. I know it's not the CPU or ram because I tried changing it, same problem.

 

[Holst1981]

Try the bios flash.

You shouldn't need to do anything special to run one of these cards.
If its not working on a fresh windows install with the correct drivers installed then it might be worth thinking about rma.

 

[Nasher]

Just tried a re-BIOS. Will see how it goes.

Tried a windows re-install as well, didn't help

 

[D1craig]

i could get crashes while just browsing the web a while back. there was a problem for a long time where some drivers where causing crashes on some systems. there was never a solution as far as we where informed, it just either sorted itself out or they fixed without telling us about it. i havent had the driver crashes i was having for a while now. maybe as long as 4 months.

i actually have 362 driver on my desktop as my fall back driver. but i think the actual driver i settled on at the time was closer to 330. cant rember what the driver was exactly though.

i never actually checked for error logs so this COULD be totally different but it sounds similar.

 

[Nasher]

I upped the voltages slightly and it SEEMS to be stable now, needs more testing though.

If that is the problem it's not the first time I've seen an under-volted nvidia card :/

 

[Davedree]

Was it throttling when getting the black screens?
When it throttles it'll reduce by 13mhz and some voltage, the voltage can then be too low for the boost clock and crash eventually.

 

[Nasher]

I think the problem is the factory has OCed the card, but not increased the voltage from standard

 

[Reality|Bites]

Set up a fan profile so that the fans spin at least a little when not gaming. The GPU temps may be fine but something else on the card may be getting a little hot and causing your problems.