Can we have an answer from all  "your house will burn soon" members ? 

yeah ACX 4.0 even if just for PR reasons needs like a Heatpipe directly touching the VRM, not just the GPU.  
 
the thermal pad solution (especially the one between the faceplate and the heatsink, which will collect DUST like crazy over time) is pretty Ghetto, but i'm not one to argue with results... next time though, Heatpipe!  

Good news .
The EVGA with thermal pads and with a custom fan curve (so acceptable Db level and heat) should now be atleast as well as others brands.
 
I'm not sure,
did they include the blackscreen issue in the 0.2 % ?
 
I should receive my 1080 FTW RMA tomorrow, I can't wait

Seb6440

ilyama
Can we have an answer from all  "your house will burn soon" members ? 

It doesnt change anything about "exploding parts" ...
They have proofs for that.
You take it funny but you wont if it happens to you.

Video card parts have been "blowing up", "burning up", and "catching on fire" for years.  This is nothing new.  I guarantee that EVERY video card from EVERY manufacturer has seen at least one example since the dawn of video card time.  According to EVGA, the defects per million trend for these cards is not significantly different than what is normal in the past.
 
Move along.  Move along.  We simply have an internet Reddit-style hype culture bringing out more examples of the 0.0001% of cards which are having the same issues other video cards have had over the last decades.

Lazy_Bones

ZeroBANG
i think it is safe to say now that EVGA should ship Tinfoil Hats instead of Thermal Pads.  
 
GG on the Internet Panic though.... 0.02% failure rate and people panic like we are looking at the next 33% RROD situation.  

 
The missing thermal pads were, in my mind, a legitimate concern to have, as they were omitted by mistake and not design.
 
Worrying about the 0.02% is just plain silly. Especially since Evga has your back if you turn out to be one of the (very) few who has a capacitor blow up, and will most likely replace the card and any parts damaged by the card.

strictly speaking this is not a mistake but a design.
using midplate/cooling plate/base plate, whatever you call it, to cool the vram and vrm is nothing "wrong" and common eg msi gamings , galax hof, inno3d ichills..... 
midplate helps reduce gpu temp because two main heat sources on pcb are now separated to two independent heat dissipation (midplate and main heatsink). 
 
the way evga is now solving the problem is to connect two independent heat dissipation... which is not the native purpose of using midplate cooling.
 
some cards are using heat direct touch method to cool the vrm and vram from the main heatsink. these main heatsink design have a unique feature to cool the vrm and vram. to maximize the contact area of main heatsink to vrm and vram, horizontal fins (90 degrees folded) must be made. because evga was not intended to use main heatsink to cool the vrm, so the acx3.0 heatsink doesnt have folded horizontal fins to cool the vrm. 
 
however actually the result of adding pad between evga midplate and evga main heatsink is surprisingly good. it shows the chokes were very hot before and using heat direct touch method is way better to cool down the vrm temp. if evga acx3.0 heaksink has folded fins for cooling vrm then there will not be a thermal issue at all. reducing vrm temp also helps in lowering gpu temp a bit.  
 
adding pads to the back of the pcb for spreading heat to the backplate is really an additional way to cool the vrm and vram. there are few companies doing this like gigabyte waterforece and xtreme. if the midplate//main heatsink is good enough to cool the vram and vrm, backplate cooling is really optional and additional.   just simply dont trust backplate can help cooling performance without pads to make contact with the back of the pcb to the backplate.
 
 
 
 


 
 
 
the evga pad added between the midpate and acx3.0heatisnk:

 

I guess this would explain why the non-FTW cards (i.e. reference cards with ACX 3.0 cooler) have been fine as far as I know.

This article already relates to so many threads that have already been created...

bobmitch
Very interesting!  Seems that thermals are quite normal, which is good to settle the mind.  I always had one question.  The ACX 3.0 and SC versions of the card use the reference pcb with one 8 pin vs all the other cards having two 8 pin.  I always wondered of the extra power to the FTW, and Classified did something to accelerate or make the issue happen.  Very rarely do you read about a reference ACX 3.0 1080 or SC "blowing up".  Just a thought...

I mentioned this repeatedly that the FTW was having this problem, but it was overlooked regarding the power phases and it was completely relevantto the issues the 980FTW suffered, having almost the same issues. Even still, with this in mind, the 1080FTW still suffered with overheating problems. I think y'all learned something; stay away from the FTWs and stick the reference models, you'll get, probably, the same overclocking capability and save a good chunk of change. Live and learn I guess.

arokhantos
Suggestion to evga replace every card and recall all bad cards, mistakes happen no big deal, annoying defiantly.

Only 0.0001% are "bad".  How can EVGA know which ones to recall until they fail?  Most likely there is nothing wrong with YOUR video card and most likely you DO NOT NEED the additional thermal pads.  Most likely you do not NEED to be here worrying or complaining.