hi,
im still on my 980 Classy and still havent decided if im gonny sell my card and buy a 1070. As there is no 1070 Classy, i tried to pull as much information about the 1070 as i could. I stumbled over a hardware-testing site here in Germany, which tested the PCBs and phase-layout of different Pascal Cards (Toms Hardware). Testing included infrared-temp-analysis of the hot PCB and i was shocked to see that the 1080FTW(in this case they tested not 1070 FTW) went up to 97 degree (even 106 degree C with raised Powertarget under heavy load und Furmark). The HOTspots on the pcb were so big that they even reached the VRAM/memory area and warmed up the memory-chips.
 
They came to the conclusion that the phase-layout is far from good, so it seems the mosfets are not cooled properly or something like said. There were other PCBs that didnt develope any hotspots and the whole PCB never got warmer than 75 C...(i remember the KFA HOF to be one of them)
 
To me this information was pretty new and i would like to ask if someone can confirm this or can show me another, maybe english testsite, where they testet in this way, sensoring the PCB-Hotspots?
 
greetings from Germany
 
PS::http://www.tomshardware.d...berichte-242137-2.html

Maybe this explains the black screen crash? :O

Well that's interesting and kind of confirms my fears about the card.

The information makes sense if you consider that many people are reporting getting the 100% fans and black screen after having used the card for a month or two, if the memory ICs are exposed to about 107c degrees all the time they will fail because they will be running out spec.
 
According to Micron GDDR5X Specs sheet for these modules found in the link below, their Operating temperature is (0°C ≤ TC ≤ +95°C) so 107c continuous exposure will kill them.
 
https://www.micron.com/~/media/documents/products/data-sheet/dram/gddr5/8gb_gddr5x_sgram_brief.pdf
 
EVGA said that they will not say the reason behind the failure so that people do not attempt to fix it themselves, what else can users fix on a card other than improve the cooling themselves? i don't think that users can go as far as desoldering and replacing components on the PCB.

Maybe the backplate is holding too much heat.

Still this is happening (should say was as its been fixed) also on the hybrid ftw card and the memory is also cooled by the water cooler with a copper plate so the bs 100% issue was not heat related.

AHowes
Maybe the backplate is holding too much heat.

Still this is happening (should say was as its been fixed) also on the hybrid ftw card and the memory is also cooled by the water cooler with a copper plate so the bs 100% issue was not heat related.

You can not tell if it is heat related or not in the Hybrid unless you thermal image it in action, it looks like the VRMs are heating up the memory module legs or surrounding PCB possibly causing solder to break or ICs to fail, water cooling the surface might not be of much help.
 
Thermal imaging of the Hybrid is needed to verify the issue.

AHowes
Maybe the backplate is holding too much heat.

Still this is happening (should say was as its been fixed) also on the hybrid ftw card and the memory is also cooled by the water cooler with a copper plate so the bs 100% issue was not heat related.

There was one guy in the black screen thread that noticed the backplate was very hot since his airflow 'sucked' according to his own words and he put a fan inside his case to just move the air over and under the card and it apparently stopped the black screens. It's only a short term fix though if you ask me because I have quite the airflow on mine (Noctua Industrials + Noctua CPU cooler).

:) my air flow was tops as well.. I even had a 140mm fan mounted at the end of the cards blowing directly at them that covered both cards in sli. Always was there with prior 980 sli cards.

Who knows.

But that's what happends when you leave the people to use there imagination insted of just informing them what exactly the issue is.

AHowes
But that's what happends when you leave the people to use there imagination insted of just informing them what exactly the issue is.

Exactly, silence is the worst product policy a manufacturer can practice towards its users, they are silent regarding this and silent regarding the Micron GDDR5 issue, we have been using our imaginations a lot in this doomed Pascal generation because no one is sharing any information with us, we are facing issues that we can only use deductive reasoning to figure out, no one can blame us for making assumptions about the products we own when transparency and information sharing is non-existent!

Hello, long time I buy EVGA for the excellent support they give their products, in this case I are defrauding much as they do not clarify or fix the problem Micron or high temperatures, hopefully soon give solution to everything and explain that but it is happening and can give order to the stores to give us back the money paid for low quality products.
Greetings.