Monday, October 24th 2016

EVGA GTX 1070/1080 Overheating Issues - Company Says Thermal Pads A Solution

After users' reports (and Tom's Hardware.de testing) of EVGA FTW 1080 and 1070 cards displaying black screen issues, and sometimes even sparking and dying altogether, even at stock voltage, the company is now moving towards fixing the issue.

Apparently, the issue stems from the absence of any thermal pads over the VRM area of the FTW line of cards, which prompts higher operating temperatures. Some users were reporting heat transfer in such quantities that even the GDDR5X memory chips on the cards were being heated at 107 ºC, significantly over their rated operating temperatures of (0°C ≤ TC ≤ +95°C).
In a blog post on EVGA's forums, a company representative issued the following message:

"The test used in the referenced review from Toms Hardware (Germany) is running under Furmark, an extreme usage case, as most overclockers know. We believe this is a good approach to have some idea about the graphics card limit, and the thermal performance under the worst case scenario. EVGA has performed a similar qualification test during the design process, at a higher ambient temperature (30C in chamber) with a thermal coupler probe directly contacting the key components and after the Toms Hardware (Germany) review, we have retested this again. The results in both tests show the temperature of PWM and memory is within the spec tolerance under the same stress test, and is working as originally designed with no issues.

With this being said, EVGA understands that lower temperatures are preferred by reviewers and customers.

During our recent testing, we have applied additional thermal pads between the backplate and the PCB and between the baseplate and the heatsink fins, with the results shown below. We will offer these optional thermal pads free of charge to EVGA owners who want to have a lower temperature. These thermal pads will be ready soon; and customers can request them on Monday, October 24th, 2016. Also, we will work with Toms Hardware to do a retest."

EVGA stresses that according to their own internal testing, no overheating issues are present in the card by using it as is (i.e., without the thermal pads), and that there won't exist any long term effects on using the cards such as they are. The company reiterates that their offer stands only inasmuch as some of EVGA's users want lower temperatures specifically on the referred cards' VRM and memory, with the usage of the thermal pads not being required to achieve normal operation. In a statement, the company has also attributed the black screen issues towards an "out of spec On Semiconductor component" that affected only 3-4% of EVGA's GTX 1080 FTW/HYBRID, with these issues "not [being] related to temperatures in any way" having been fixed as of September 1st, with the company having already sent replacements where applicable.

The thermal pads will be provided to anyone with a GTX 1080 or GTX 1070 with an ACX cooler and a backplate, with other graphics card models from EVGA's models also being considered for the offer. While EVGA is surely going about the problem in the correct manner, it still leaves to customers the responsibility of dismantling their EVGA cards and applying the thermal pads on their own. Bellow is an image posted by EVGA showing the temperature results of a GTX 1080 FTW with the thermal pad solution.
Source: EVGA Blogs
Add your own comment

82 Comments on EVGA GTX 1070/1080 Overheating Issues - Company Says Thermal Pads A Solution

#26
jormungand
The more i read posts here the more i cry :( i think i will have to cross my fingers and hope my card doesnt fu.... And i was thinking on grabbing the msi gaming z b t decided for evga cuz was evga of course.... And ended in a curse
Posted on Reply
#27
TheLaughingMan
jormungandThe more i read posts here the more i cry :( i think i will have to cross my fingers and hope my card doesnt fu.... And i was thinking on grabbing the msi gaming z b t decided for evga cuz was evga of course.... And ended in a curse
If it makes you feel any better, The EVGA 1070 SC was the one Wiz reviewed and the pictures clearing show thermal pads everywhere. Hopefully this was a line manufacturing issue with the FTW batches only.
Posted on Reply
#28
GC_PaNzerFIN
Ferrum MasterThe shown burnt cap usually isn't caused by temperature but defect and leak, thus it shorts out, wrong markings lover real voltage ratios.

It could be totally digging the wrong way.
Most SMD caps are rated at 85 degrees C, the good stuff is rated at 105C. What do you think running them past their maximum rating would do other than cause defects, increase leakage current, finally short out and burn?
Combined effect of both voltage and temperature derating and there you go. Marginally OK caps go boom sooner, rest later. Heat kills caps. Not just the electrolytic kind.
Posted on Reply
#29
Ferrum Master
GC_PaNzerFINMost SMD caps are rated at 85 degrees C, the good stuff is rated at 105C. What do you think running them past their maximum rating would do other than cause defects, increase leakage current, finally short out and burn?
Combined effect of both voltage and temperature derating and there you go. Marginally OK caps go boom sooner, rest later. Heat kills caps. Not just the electrolytic kind.
This particular one in the pic looked more like a dud. Actually there are no problems getting ceramic caps that hold +200C, usual ones are +125C... those are not lytics exactly...
Posted on Reply
#30
Supercrit
Ferrum MasterThis particular one in the pic looked more like a dud. Actually there are no problems getting ceramic caps that hold +200C, usual ones are +125C... those are not lytics exactly...
Does it matter if they can survive 1000C if other components nearby melted due to high temps? It's just bad, no need to discuss further.
Posted on Reply
#31
Ferrum Master
SupercritDoes it matter if they can survive 1000C if other components nearby melted due to high temps? It's just bad, no need to discuss further.
Could you be reasonable and READ??

They should operate normally... I suspect a bad batch(or a human error) of parts not some missing tapes or bad thermals(like putting a patch to a corpse). 100C in VRM section isn't something uncommon(R9 290)... not that have died much from that being a fire spitting monster and being constantly farming in bitcoins...
Posted on Reply
#32
jormungand
TheLaughingManIf it makes you feel any better, The EVGA 1070 SC was the one Wiz reviewed and the pictures clearing show thermal pads everywhere. Hopefully this was a line manufacturing issue with the FTW batches only.
And i have the ftw only.....
Posted on Reply
#33
$ReaPeR$
i thought evga had better quality standards.. pitty.
Posted on Reply
#34
Supercrit
Ferrum MasterCould you be reasonable and READ??

They should operate normally... I suspect a bad batch(or a human error) of parts not some missing tapes or bad thermals(like putting a patch to a corpse). 100C in VRM section isn't something uncommon(R9 290)... not that have died much from that being a fire spitting monster and being constantly farming in bitcoins...
The thing is the VRM might not care about the high temps, but not the components nearby. High temp in anything is bad in general, especially when the memory is being cooked.
Posted on Reply
#35
Dave65
You really have to wonder, who's bright idea was this, whoever it was he/she should be fired.

EVGA should be replacing these cards with a brand new unit..
Posted on Reply
#36
GhostRyder
That is unfortunate and unacceptable for a premium version of the card to have that issue especially from EVGA no less. I mean, its not the end of the world, but for a high overclocking card (At least thats what they used to intend) I expect the cooling to be top notch.

Again, an easy fix but it seems like something that should have been avoided to begin with.
Posted on Reply
#37
jormungand
GhostRyderThat is unfortunate and unacceptable for a premium version of the card to have that issue especially from EVGA no less. I mean, its not the end of the world, but for a high overclocking card (At least thats what they used to intend) I expect the cooling to be top notch.

Again, an easy fix but it seems like something that should have been avoided to begin with.
i dont think is an easy fix since you may lose your warranty on the card.... for me was $459 at newegg so my opinion on EVGA was destroyed by themselves. Since ive had a line of good cards (gtx 570 SC-670 -970 SSC)
Posted on Reply
#38
n-ster
EVGA is still a good brand despite their fu ... This seems to be only on their FTW line and they are admitting their fault and trying to give users a solution. Better yet , you know that kind of issue will be absent in the future as there will be more scrutiny around that etc.

Around here, they are usually the best bang per buck cards. I can safely say I'll still be recommending EVGA cards

Seems paranoid to think that they will void warranties for people trying to install the supplied thermal pads.
Posted on Reply
#39
EarthDog
I mean... stuff happens...

To put them on a public trial like I see some are doing for a simple mistake (which they are correcting), is more disconcerting than EVGA's mistake here.
Posted on Reply
#40
Legacy-ZA
*Sigh* After getting a screwed up MSi card; this was the last thing I wanted to read after I swapped to the EVGA FTW. :*(
Posted on Reply
#41
Basard
So, no replacement cards? F that! Who knows what kinda damage has already been done to the card.

Here, burn your card up for 6 months then dismantle the entire thing and put these thermal pads on. It's on the house.

Where's everybody's $75 an hour for labor? After all you have to be an "expert" to take your card apart without voiding your warranty, right?

I feel bad for you guys that bought one.
Posted on Reply
#42
bug
GhostRyderThat is unfortunate and unacceptable for a premium version of the card to have that issue especially from EVGA no less. I mean, its not the end of the world, but for a high overclocking card (At least thats what they used to intend) I expect the cooling to be top notch.

Again, an easy fix but it seems like something that should have been avoided to begin with.
Frankly speaking, this should kill the FTW brand like exploding batteries killed Samsung's Note.

Also, anyone still remember s how EVGA screwed Unwinder over RivaTuner code? It seems the brand has been going downhill for a while. I've just got my card from them, but I'll be sure to look at alternatives first for my next upgrade.
Posted on Reply
#43
cdawall
where the hell are my stars
BasardSo, no replacement cards? F that! Who knows what kinda damage has already been done to the card.

Here, burn your card up for 6 months then dismantle the entire thing and put these thermal pads on. It's on the house.

Where's everybody's $75 an hour for labor? After all you have to be an "expert" to take your card apart without voiding your warranty, right?

I feel bad for you guys that bought one.
$75 an hour to do menial labor? Where do I sign up. I also bet these cards are not failing under normal use, the failures reported are probably the batch that thinks they need to run furmark on everything for no reason.
Posted on Reply
#44
R-T-B
This is piss poor on EVGA's part. I'd expect a free replacement offer/exchange program from EVGA.

It's sad, because I had a really good RMA experience from EVGA's mobo department. But then again, they did blow up half my rig with a defective card. :P
Posted on Reply
#45
drewskidrewsgv
It's funny how people don't understand that these "tests" by the German Tech Power up site were done using Furmark for AN HOUR AND A HALF. That's why those temps got so high and also Furmark isn't even supported by Pascal and Maxwell GPU's. EVGA didn't even say this was a solution, they just did this to stop the whining from people about a non existent issue.
Posted on Reply
#46
alucasa
Me stopped reading when me spotted "furmark".

That thing is used only when a user wants a replacement card, periold.
Posted on Reply
#47
Jism
drewskidrewsgvIt's funny how people don't understand that these "tests" by the German Tech Power up site were done using Furmark for AN HOUR AND A HALF. That's why those temps got so high and also Furmark isn't even supported by Pascal and Maxwell GPU's. EVGA didn't even say this was a solution, they just did this to stop the whining from people about a non existent issue.
Yeah, furmarking your card to pieces is the way to go.

It puts an unrealistic load to your VRM's, GPU and Memory, causing the thing to heat up. It's not bin tested for QC-standards like this.
Posted on Reply
#48
msamelis
I simply cannot trust EVGA again. Remember when they did a similar thing with the ACX 2.0 and ACX coolers on the 7 and 9 series from Nvidia and they claimed it was deliberate that only 2 of the 3 heatptipes touched the GPU core? The third heatpipe was not even sealed off.

And now this? You can say that Furmark is not realistic and all that but the rest of the competition still made sure they'd provide a proper cooler, even for unrealistic scenarios.

TPU 970 pic:


TPU 760 pic:


Notice the third heatpipe:


Shame on you EVGA.
Posted on Reply
#49
Darkleoco
Anyone besides me have a 1070 SC they could conveniently check or am I gonna end up being the only one with one :(
Posted on Reply
#50
cdawall
where the hell are my stars
msamelisI simply cannot trust EVGA again. Remember when they did a similar thing with the ACX 2.0 and ACX coolers on the 7 and 9 series from Nvidia and they claimed it was deliberate that only 2 of the 3 heatptipes touched the GPU core? The third heatpipe was not even sealed off.

And now this? You can say that Furmark is not realistic and all that but the rest of the competition still made sure they'd provide a proper cooler, even for unrealistic scenarios.

TPU 970 pic:


TPU 760 pic:


Notice the third heatpipe:


Shame on you EVGA.
To be fair the third heatpipe is a dinky little thing and it looks completely intentional.
Posted on Reply
Add your own comment
May 8th, 2024 19:50 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts