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

[cdawall]

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.

To be fair the third heatpipe is a dinky little thing and it looks completely intentional.

 

[buggalugs]

This is hilarious. Problem discovered with card. Company gives out thermal pads so user can fix it.

 

[ViperXTR]

and all of them folks keeps hailing EVGA as the grandmaster manufacturer of graphics card :V

 

[Ferrum Master]

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.

You still lack the ability to read.

This news post consist of blown capacitor element on the board. I call that rubbish and and a single accident. Such serious damage as blown single element occur rarely... Often from overvoltage(bad PSU spike). That is not the cause for black screens and is completely a different thing thrown into this article and you mindlesly chew it on without considering facts.

The vram overheating especially using furmark? That's also something else.

This looks like some speculation to gather clicks, the real issue is somewhere else.

 

[ArchStupid]

Luckily the MSI GTX 1080 Gaming X was available in Israel slightly before this, so I got that instead.
I always wondered if I should've waited a couple of more days; now I know the answer.

 

[RejZoR]

and all of them folks keeps hailing EVGA as the grandmaster manufacturer of graphics card :V

I never liked EVGA. In the past it was praised because of a beefier PCB which was favorable among hardcore overlcockers iirc, but that was back in GeForce 4 era. These days they seem just sloppy. And that they leave the burden of repair on the user, c'mon!?

 

[Melvis]

Never liked EVGA ever since Skt 775.

 

[Vayra86]

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...

You're not on the money here.

The issue arises not directly with the VRM temps, but the temps of the two GDDR5(X) modules next to the VRM section. Those are rated for 95 C and exceed that temperature. This results in black screen/artifacting even at stock.

The reported issues arise even at stock clocks for those who have issues. The only confirmed issues with hot GDDR5X are on the FTW 1080. The 1070 has a similar VRM section and cooler design, but a lower TDP.

I'm still considering whether or not I'll return my card. I've sent EVGA an RMA request to have them apply the solution themselves. I'm not burning myself on this one, I can still return the card to the seller. Even if my card runs a lot cooler than the ones in the reports... I've never seen temps go past 65 C in game, with 122% PT and maxed out OC, but still I am almost convinced the lifespan of the card is lower than it should be.

 

[Ferrum Master]

The issue arises not directly with the VRM temps, but the temps of the two GDDR5(X) modules next to the VRM section

That's the point really. The blown up cap case should not be mixed together.

 

[Vayra86]

That's the point really. The blown up cap case should not be mixed together.

Unfortunately, with EVGA's responses on the issue thus far there is no guarantee the two are not related and it seems to be more likely that they are both a result of incapable cooling in that area.

The lack of a clear response from EVGA prompted my RMA 'test' request. If they are unwilling to apply the correct solution to my card, I'm sending it back. Not taking any risk with this heavily overpriced piece of kit - even if I have an 'above average' core OC

Shame really. I hate this hassle.

 

[Basard]

$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.

People get their warranties revoked for this "menial" labor. $75/hr is what a computer tech would charge you to do something like that, if you're lucky.

People bought these cards expecting that they were fully furnished. That's what the reviews led them to believe. If they were going to "normally use" these cards they would have just bought a dell with a "4Gb Nvidia graphics card" inside.

 

[Vayra86]

UPDATE: sending card back within my return period for a full refund from seller, just got it approved.

EVGA has yet to give a clear statement and instructions regarding the offered solution and they tell me my card 'has no issues'. On my replied question as to the impact on longevity of the card, they responded by suggesting I'd better buy into their 5 year extended warranty plan. I'm not even going to dignify that with a response because I just can't be arsed to - EVGA customer service is on the ball, but clearly this fuckup was concluded to cost them waaay too much money because in essence, EVERY card is affected and they want to keep that fact as low-key as possible. Even though from a business perspective I fully understand this, for me personally it's something I do not accept. The vast majority of users won't be pushing 99% utilization with an OC from day one, but increased degradation is the only possible result of this cooling design. This 'fact' was not confirmed or denied - in other words, nuff said.

The fact they try to push me into extended warranty says enough. That's as much as implicitly stating it will break down eventually, and most likely just after 3 years have passed.

The bottom line of this story is: avoid the FTW 10xx or demand that it comes with the thermal pad fix (EVGA confirmed they will incorporate that into the manufacturing process now) AND SAMSUNG VRAM. The Micron memory clocks consistently lower. Lesson learned ^^

 

[TheLostSwede]

This is the URL to get your strip of thermal tape: http://www.evga.com/thermalmod/

 

[Frick]

I'm just sad more cards haven't spontaneously combusted.

 

[cdawall]

People get their warranties revoked for this "menial" labor. $75/hr is what a computer tech would charge you to do something like that, if you're lucky.

People bought these cards expecting that they were fully furnished. That's what the reviews led them to believe. If they were going to "normally use" these cards they would have just bought a dell with a "4Gb Nvidia graphics card" inside.

Where are you that pc techs that do basic pc work make $75/hr? I need to move.

 

[64K]

Where are you that pc techs that do basic pc work make $75/hr? I need to move.

He's probably talking about what a PC repair shop charges per hour but I knew a guy that worked at a repair shop and he probably would have done it for free since it would only take a couple of minutes. If the repair was big enough to take the computer in then there was a minimum 1 hour charge.

 

[cdawall]

He's probably talking about what a PC repair shop charges per hour but I knew a guy that worked at a repair shop and he probably would have done it for free since it would only take a couple of minutes. If the repair was big enough to take the computer in then there was a minimum 1 hour charge.

That's pretty typical where I work and the shop would have installed any and all of those kits for free if the card was purchased at the store.

This also still isn't a real problem the faster people realize furmark is dumb the better. Let it kill their damn cards. Consider it darwinism

 

[Vayra86]

That's pretty typical where I work and the shop would have installed any and all of those kits for free if the card was purchased at the store.

This also still isn't a real problem the faster people realize furmark is dumb the better. Let it kill their damn cards. Consider it darwinism

Metro : LL at 4K was also running into the 95 C temperature limit for the GDDR5x and showed on FLIR as being as hot as 97 C. Also, note that this is on a freshly unboxed, OPEN CASE, super clean GPU and cooling block with fresh paste. Not one that is running for months or a year in a normal environment and closed case.

Furmark simply stabilizes the power draw at max while benches and games are more spiky in behaviour. Nothing else. Furmark is therefore STILL a good stress bench, and will immediately show shoddy cooling instead of normal benchmarks that need many runs to give you a glimpse. Even EVGA agreed on that, btw.

That's darwin for ya right there... I would say apply Darwinism to the FTW cooling solution, and the only conclusion is that it cannot keep up with the rest of the product and needs replacing. ACX 3.0 lost the race.

 

[cdawall]

If Furmark is as good as you say it is why does both amd and nvidias driver specifically block it from running at full clocks?

It is a power leach nothing compares even folding or bitcoin mining on the cards.

EVGA is providing a solution. If you don't like their solution don't buy their product. They are however the number one selling cards in the states...

Another note open test benches aren't the best for everything. Unless you have a GPU fan specifically blowing across the backplate of the card most standard cases would keep the RAM and VRM cooler.

 

[hyp36rmax]

after sapphire, evga also have crapped their cards and that to on FTW. they must be cheapskates not to put on thermal pads that cost maybe 1-2$ max.

Don't think it has anything to do with being cheap. Could be an oversight which has prompted EVGA with a quick solution. You seem to have a chip on your shoulder...

 

[jormungand]

Look up on newegg reviews and still dont see none complaining about this but who knows when is gonna die.......

 

[cdawall]

Look up on newegg reviews and still dont see none complaining about this but who knows when is gonna die.......

It normal system usage? Probably never. Overclocked+running furmark+0 airflow, pretty quickly from the sounds of it.

 

[Prima.Vera]

and all of them folks keeps hailing EVGA as the grandmaster manufacturer of graphics card :V

I though MSI was leading the pack...until they start to ridiculously overprice all of their cards.

 

[Vayra86]

If Furmark is as good as you say it is why does both amd and nvidias driver specifically block it from running at full clocks?

It is a power leach nothing compares even folding or bitcoin mining on the cards.

EVGA is providing a solution. If you don't like their solution don't buy their product. They are however the number one selling cards in the states...

Another note open test benches aren't the best for everything. Unless you have a GPU fan specifically blowing across the backplate of the card most standard cases would keep the RAM and VRM cooler.

You keep going on about Furmark but I've already provided you with a similar example of in-game temps (and those are just 2-3 C lower than in Furmark). You seem very convinced about your own truth, but the fact is, EVGA is even applying the thermal pads to all future product on the ACX 3.0 coolers *and* provides them to everyone who has one. Some perspective... Read between the lines here. They chose this solution *because* they were getting an awfully high number of RMA's on the FTW cards, especially the 1080. Just two days ago they announced all RMA procedures were on immediate hold (management decision) and a day after that we get thermal pads.

The PSA regarding those thermal pads is a nice little spin on the issue which gives you the idea they are being very nice to all FTW customers, but this is damage control and nothing else. You even get to apply them yourself, how awesome is that?! /sarcasm

 

[EarthDog]

Furmark simply stabilizes the power draw at max while benches and games are more spiky in behaviour. Nothing else. Furmark is therefore STILL a good stress bench, and will immediately show shoddy cooling instead of normal benchmarks that need many runs to give you a glimpse. Even EVGA agreed on that, btw.

No, it isn't... it really isn't at all IMO. As I said, perhaps at another site, what is the point of using Furmark when it throttles at stock speeds? When I game, yes, even in Metro:LL, the boost bin stays where I set it. When I furmark at the same settings, I lose several boost bins. How is that good to use when you aren't even testing the clocks you intend to test????

How many cards have 'shoddy cooling' to being with? I mean most have 'shoddy cooling' when you run Furmark........... oh wait, it throttles modern cards back at stock speeds... my fault.........