EVGA 2080 Ti Black 0.5mm Gap Issue...

[integr8d]

Curious as to if anyone from AT proper would be interested in taking this on...

Confirmed. EVGA has a major gap between the silicon and the cold plate. So far, they’re not acknowledging it.

 

[Guru]

How does 1 video with 1 card prove there is an issue with Evga 2080ti black cards?

 

[integr8d]

1 video card, 1 stock air cooler and 1 Hydro Copper water block... I would imagine the z-height for the silicon is pretty consistent from one board to the next. And the air cooler and water block both failed to make contact w/ the silicon. As to your point, that's why I'm trying to get media attention; A: to confirm my findings and B: to be the voice for people spending generous amounts of money for GPUs, if the findings are correct and consistent.

 

[SilentRunning]

The guy who made the video showed that he should leave these projects to others. Just watching him smear compound on contacts around GPU when removing it makes me cringe. Most obvious check would be to remove mosfet/vrm thermal pads and see what contact is like, the looked to be doubled up. Oh and lets not forget his application of mixed compounds at the end is wrong on so many levels.

 

[thesmokingman]

Hasn't this happened before in the past with evga? That is stack heights...

 

[integr8d]

The guy who made the video showed that he should leave these projects to others. Just watching him smear compound on contacts around GPU when removing it makes me cringe. Most obvious check would be to remove mosfet/vrm thermal pads and see what contact is like, the looked to be doubled up. Oh and lets not forget his application of mixed compounds at the end is wrong on so many levels.

If that made you cringe, you'd pee yourself if you saw me delid my old 4790k...

 

[Stuka87]

If that made you cringe, you'd pee yourself if you saw me delid my old 4790k...

Why did you delid a Devil's Canyon? The 4690K and 4790K had none of the heat spreader issues early Haswell chips had.

EDIT: Oh, and as for the video, I have zero trust in what the guy is saying/showing. If somebody reputable that knows they are doing talks about this, ok. I had to chuckle at him thinking paper towels were a safe thing to work on.

 

[integr8d]

Why did you delid a Devil's Canyon? The 4690K and 4790K had none of the heat spreader issues early Haswell chips had.

EDIT: Oh, and as for the video, I have zero trust in what the guy is saying/showing. If somebody reputable that knows they are doing talks about this, ok. I had to chuckle at him thinking paper towels were a safe thing to work on.

Probably thinking of 4770 then... It's funny. 90% of the comments I get are from people who attack the messenger. Total learning lesson for me And I don't need your trust. All I need you to do is put your hands together and understand that that's how the GPU silicon and cold plate are supposed to interface w/ one another. And then I need you to pull your hands 1" apart and understand that that's how the silicon appears to be (not) interfacing w/ both the stock air cooler and the EVGA-branded water block...

PS: First PC build was an Athlon XP in 2002. I've never had an issue with using paper towels as a base. And I even try to get as much static electricity charged into them as possible...

 

[NeonFlak]

Used two different TIMs? That should be in the video description so people don't waste their time on this.

 

[integr8d]

It's a waste of time to know that the die and factory air cooler and factory water block aren't mating properly b/c someone royally screwed up on their measurements? FYI, you're the second person to make this statement and so far I've seen ZERO difference between the temps of the 'stock' TIM and TIM solution that I had to use.

0.5mm copper plate incoming and will be machined to the die size of the chip. That we'll be the primary interface to make up for the gap. Stress testing will be conducted with standard TIM, liquid metal and then finally on the water block (with LM). I'm anticipating a 10 degree drop on the air cooler and much more on water. Then people will know what they have to do to get the most out of the non-binned, entry-level Ti's.

 

[beginner99]

Even rubber can only get so thin and the screws have springs, so that you don't over tighten. This gap simply is how thin the rubber can get with the pressure applied. If you remove the rubber, the screws could be tighten a bit more and hence no gap. Why use that rubber thing? Just apply small amount paste in the middle and see how it spreads.If the gap is that big there would only be small area of paste in the middle.

EDIT:
lol and the amount of paste.

 

[ericlp]

Well, the proof would be in the temps. I didn't watch the entire 44 minute vid... Tho, if for some reason his card is getting different temps... then maybe he got a bad one? He should buy 10 more video cards, from EVGA and do the test on all 10, to see if his is bad! LOL!

Harsh reality is, I am sure EVGA is just slapping these things together... The bottom line is MONEY, if you want a better card with better cooling (copper cold plates) he should step it up, otherwise get in line! You are just buying the "good enough" version AKA consumer grade. If your thinking EVGA is gonna go out and hand pick and check every cold plate for high tolerances ...Guess again... I'm thinking the 2000 dollar kingpin cards have better tolerances. Again, it comes down to MONEY and if you want the cheap aluminum plate, or the copper one.

Don't want to step it up? That's fine, lube it up more with a bit more thermal paste and game on!