Wondering if anyone has had any recent experience with this issue. I'm kind of scratching my head at the moment because it only happens in certain games or pushing the GPU with 3D Mark. I'm leaning towards it potentially being power related but I'd have figured an 850W PSU (EVGA SuperNOVA 850 G3) would be enough.
 
While playing the game or the  benchmark typically within a few minutes you'll hear the fans ramp up immensely and then the screen goes black. You can still hear audio/system sounds but the PC is unresponsive to anything so you reboot.  The only way I've been able to get this stable is to to set the power limit to 85-90 in after burner.   Temp wise the card idles around 29-33 and in games it'll jump to 60-65 so I don't think its an overheating issue. 
 
I've tried clean installs, uninstall/reinstalling the games that create the issue but nothing else works besides the power limit.   I've seen some older forums posts/reddits with similiar issues and it seems to be a mixed bag but quite a few people saw improvements after going to 1k Watts or more, is that really the only way here? 

rjohnson11
The first thing to try is to obtain a certified DP or HDMI cable and retest.

Yup already using certified DP cables, tried HDMI as well, no changes in the black screen fan ramp scenario. 

Just had my 1000W EVGA GT show up today and  sadly the crashing still happens. the card isn't overclocked or anything with MSI afterburner but the only way to not have it happen is to adjust the curve or set the power limit to 85-90%. 

HeavyHemi

Stephenk291
Just had my 1000W EVGA GT show up today and  sadly the crashing still happens. the card isn't overclocked or anything with MSI afterburner but the only way to not have it happen is to adjust the curve or set the power limit to 85-90%. 

Generally speaking, when reducing the GPU clocks increases stability, the culprit is the GPU. The other common cause is the PSU. As you've already replaced the PSU, you're down to the GPU being faulty.

Yup that's where I'm leaning, just opened a support case.  I'm curious how the RMA works for a card being modified with the hybrid cooler, in terms of do I have to put the stock one back on, ship it as it. Guess ill find out. Kind of a bummer but at least with Cross ship I won't be dead in the water. 

HeavyHemi

Stephenk291

HeavyHemi

Stephenk291
Just had my 1000W EVGA GT show up today and  sadly the crashing still happens. the card isn't overclocked or anything with MSI afterburner but the only way to not have it happen is to adjust the curve or set the power limit to 85-90%. 

Generally speaking, when reducing the GPU clocks increases stability, the culprit is the GPU. The other common cause is the PSU. As you've already replaced the PSU, you're down to the GPU being faulty.

Yup that's where I'm leaning, just opened a support case.  I'm curious how the RMA works for a card being modified with the hybrid cooler, in terms of do I have to put the stock one back on, ship it as it. Guess ill find out. Kind of a bummer but at least with Cross ship I won't be dead in the water. 

Something else to check since you added the hybrid are your temps. Any high temps show up monitoring with GPU-Z? Yes you must put the stock cooler back on and use original thermal putty on the VRMS. Contact EVGA to purchase what you need. They've become a real stickler for the correct pads and thermal putty being used for RMA returns.

No high temps, usually tops out at  65ish with the hybrid cooler on. That's a bit of a bummer about the return, having to buy additional stuff to facilitate an RMA for a card that is defective in some manner. Guess ill get an official reply soon from the ticket. 

lascivious1
I have a 3080 ftw3 12gb. I get random black screens. But, without the crashes. I've tried everything I can think of to get rid of it. I'm running a1000w psu. Only thing I notice that fixes it, is pulling the hdmi cable from the TV and reseating it and problem is resolved till the next time I reboot. You think I am facing the same problem of the gpu?

Have you tried a different HDMI cable? If you're running 4K120hz 10 bit RGB 4:4:4, you need a cable rated for 48Gbps.

lascivious1

HeavyHemi

lascivious1
I have a 3080 ftw3 12gb. I get random black screens. But, without the crashes. I've tried everything I can think of to get rid of it. I'm running a1000w psu. Only thing I notice that fixes it, is pulling the hdmi cable from the TV and reseating it and problem is resolved till the next time I reboot. You think I am facing the same problem of the gpu?

Have you tried a different HDMI cable? If you're running 4K120hz 10 bit RGB 4:4:4, you need a cable rated for 48Gbps.

Yes, it's 48gb 2.1. Says its rated for 8k at 60hz and 4k at 120hz I'm running a 4k 120hz Sony X85k tv.

Are there any firmware updates for your set? If you run at 4K60hz does the issue go away? For testing have you set Prefer Max Performance in the NVCP? (This requires a reboot to take effect)

Does anybody have any good recommendations for thermal pads or putty. Apparently per support my RMA is going to be denied If i swap the original cooler back on and the pads and grease are not as they'd be from the factory. It was then suggested I can pay for pads and shipping to place back on the stock cooler before I submit the RMA or I can just eat a $45 charge.  Frankly after spending nearly $1500 on the card and another $150 on the cooler that actually kind of pisses me off.

Sajin
No original thermal paste will get rma denied? Wow. That is new.

Unless something was getting lost in the conversation I said If I put the stock cooler back on the the original thermal putty is sort half there half not since some came off during the removal to put the hybrid cooler on.  I said can I just put the stock cooler back on knowing some of this wont be like it came from the factory, the tech support guy said no, but I could buy pads and grease or pay some $45 charge?  I suggested maybe as an alternative could I just ship back everything and just get a hybrid replacement to avoid the hassle which was shot down immediately.  Its kind of bewildering to me I'm going to now have to spend additional money to ship back a defective card, its not the dollar amount its the principle of the matter.  So now once I remove the hybrid cooler I'm going to have to buy new pads, putty, whatever, to replace all that as well. 
 
Been buying EVGA products since the 8600GT back in the day and its kind of making me rethink the brand loyalty after that phone call.  Customer service seems to be quite a bit different now. 

Pad set is like 10 $ and better than trying to see if they still do a 45$ repad fee and chance denial , as for putty , do you not think you have enough to cover , if the pad is ripped it should be ok , not much they can say about that, just do not want it to be missing, I wrap my cooler in plastic wrap after removal with all the pads on it to store it.

HeavyHemi

Stephenk291

Sajin
No original thermal paste will get rma denied? Wow. That is new.

Unless something was getting lost in the conversation I said If I put the stock cooler back on the the original thermal putty is sort half there half not since some came off during the removal to put the hybrid cooler on.  I said can I just put the stock cooler back on knowing some of this wont be like it came from the factory, the tech support guy said no, but I could buy pads and grease or pay some $45 charge?  I suggested maybe as an alternative could I just ship back everything and just get a hybrid replacement to avoid the hassle which was shot down immediately.  Its kind of bewildering to me I'm going to now have to spend additional money to ship back a defective card, its not the dollar amount its the principle of the matter.  So now once I remove the hybrid cooler I'm going to have to buy new pads, putty, whatever, to replace all that as well. 
 
Been buying EVGA products since the 8600GT back in the day and its kind of making me rethink the brand loyalty after that phone call.  Customer service seems to be quite a bit different now. 

Clarify this with EVGA. I'm pretty sure pretty much any standard TIM between the GPU die and cooler/hybrid unit is okay. What they are being a stickler on, is the thermal putty and thermal pads.

I've emailed them to ask about the replacement pad set. this is the original cooler that was removed from the card. almost all the putty still remains on the cooler itself, I'm curious if this would suffice once I add TIM back on. 
 
EDIT: EVGA RMA manager approved the current putty that remains on the card for the RMA to be sent in and ill just be replacing the TIM.  Ordered $10 worth of replacement pads as well.