Today I've started getting "nvlddmkm.sys" BSoDs.
I've tried completely uninstalling all GPU drivers and reinstalling but I've gotten another BSoD shortly after trying that.
 
Could this be related to any of these issues in any way? I assume it's not but at this point I just don't know anymore.

tw1l1ght
I have the 1080 ftw. I am a bit confused here. I have the thermal pads being sent to me, but i did not install the bios. On the link there are two of them, primary and secondary. Do i download both of them or secondary? Also do i really need to get the bios update, i set the fan curve to go to 50% when my card goes above 45c. My temps never hit 60c. SO what do i download primary then secondary, or just the secondary one? Also is this safe, can anything go wrong. I heard of bricking a card can i do that when installing the new bios. Is there a switch on the GPU the 1080 ftw acx 3.0 that can restore the bios?

If you don't use the secondary bios you can only download the primary and flash that. No, if you're already using a custom fan curve then you are already doing what the new bios does. The new bios only changes the fan curve to be slightly more aggressive.
 
I'ts actually pretty safe, yes you have the chance of bricking your card but you also have a dual bios. So if something goes wrong whilst you're flashing the primary bios you can always flip it over to the secondary and fix it. It literally takes about 10 seconds from start to finish (not even) and unless you downloaded the incorrect bios you'll be fine. It takes longer and is more complicated to flash a motherboard bios, that's how easy it is to do it via EVGA's method.

Scorpion-X
Got my replacement cards on Wednesday, temps are much better overall, originally maxed about 73-75 degrees, now 63-65 degrees (both cards set to 80% fan speed in 3D mode) and (yes I know the thermal pads are for the VRM's but it looks like the heat was transferring to the GPU because of how hot it was, also back of card seems a bit cooler also, both sets of cards had the latest BIOS update) but the overclock is not very good, with my previous cards I was getting 2100+, now I'm getting at best 2025Mhz, and even at that I still see a random artifact once and awhile, I also tried swapping the cards, made it worse, couldn't get anything over 2000Mhz without it crashing....is this acceptable? What do you guys think?

 
Probably, a lot of people with lower-overclocking cards see the EVGA thermal pad installation as a chance to ditch their slower cards and maybe get a faster ones, in return. While people with higher-overclocking cards are not sending theirs in as mcuh, because they don't want to lose their good overclocks, and so they're installing thermal pads, themselves. The result of this probable situation would be that there's a disproportionate amount of lower-overclocking cards being sent to EVGA, and sent back to people after EVGA installs thermal pads on them.
 
Also, if the cards being sent to EVGA already have heat damage, they might perform less than they first did. EVGA will still install thermal pads, check to see that the cards perform at their advertised speed, and then send them out again.
 

tw1l1ght
Awsome thank you for your help. Can you also please explain what exactly i have to do to fix the card if i brick it?

If you brick it, then you have to RMA it to EVGA, to get a replacement.

Triss

ilyama

Triss
Any customer who is not comfortable performing the recommended VBIOS update, may request a warranty cross-shipment* to exchange the product to EVGA for an updated replacement.
 
Since being from New Zealand i'm not in the listed countries for advanced RMA  i presume i have to first send in my gpu and wait for a new one to be sent back to me?
I will say if that is the case this will be my last evga product sadly.

 
Just change the fan curve... 

Changing the fan curve will not also add thermal pads....
I'm already going to end up with a louder product then was reviewed and promoted i accept that but i wont be doing all these changes myself.  
This is why EVGA has offered the exchange policy.
 

There is really nothing to worrie about. If your card is working fine, (and most cards do) playing games, surf, do work, you don't have to apply pads or update VBios.
 

DrPhD
Today I've started getting "nvlddmkm.sys" BSoDs.
I've tried completely uninstalling all GPU drivers and reinstalling but I've gotten another BSoD shortly after trying that.
 
Could this be related to any of these issues in any way? I assume it's not but at this point I just don't know anymore.

Do you overclock?

Anarion

emsir

delicieuxz

emsir

Anarion

clone
I must be one of the lucky ones as my 1070FTW runs very cool but ive still asked for the thermal pads.

GPU might run very cool but the VRM are could be well past the boiling point.

This is rubbish, you have no documentation to prove your statement. Don't post something you don't have a clue about.

What do you find to be inaccurate about their statement? The temperature sensor on graphics cards is for the GPU chip, and not the VRM or VRAM. Whatever temperature is read from the GPU chip does not reflect the thermal conditions of the VRAM and VRM.
 
I hope that a cool GPU means also cool VRM and VRAM, but without thermal pads to transfer the heat from them, there seems to be significant risk to those components.

I've had my GTX 1070 FTW since September and played a lot of demanding games. No issues at all. So significant risk to components are not true. There haven't been 1 card with damaged VRM or VRAM. Nothing! And I think a lot of people with EVGA 10 series card are using agressive fan curve in Presicion X, and by that the card has no heat issue at all.  

LOL. That doesn't mean your VRM are wouldn't be running at >100c. When VRM fails, it usually fails without any warnings. Massive majority just installs the card and doesn't install Presicion X let alone use custom fan curve. I don't use Presicion X, I use Afterburner and it doesn't have any preset fancurves. Besides, the aggressive one in Presicion X is loud as hell judging by screenshots that people post.
 
The worse the case cooling, the bigger the problem. Especially if you do not do anything to the card. If everything was completely fine EVGA wouldn't kindly ship those thermal pads and update BIOSes.

If the VRM runs at 100 degrees  it will for sure have an effect on the GPU temps. If something gets really hot it will affect the surrounding areas = GPU!  About your argument about if VRM fails due to overheat. No (EVGA) card have failed due to overheated VRM. If you have some info about cards bricked, please post it. 

emsir

Anarion

emsir

delicieuxz

emsir

Anarion

clone
I must be one of the lucky ones as my 1070FTW runs very cool but ive still asked for the thermal pads.

GPU might run very cool but the VRM are could be well past the boiling point.

This is rubbish, you have no documentation to prove your statement. Don't post something you don't have a clue about.

What do you find to be inaccurate about their statement? The temperature sensor on graphics cards is for the GPU chip, and not the VRM or VRAM. Whatever temperature is read from the GPU chip does not reflect the thermal conditions of the VRAM and VRM.
 
I hope that a cool GPU means also cool VRM and VRAM, but without thermal pads to transfer the heat from them, there seems to be significant risk to those components.

I've had my GTX 1070 FTW since September and played a lot of demanding games. No issues at all. So significant risk to components are not true. There haven't been 1 card with damaged VRM or VRAM. Nothing! And I think a lot of people with EVGA 10 series card are using agressive fan curve in Presicion X, and by that the card has no heat issue at all.  

LOL. That doesn't mean your VRM are wouldn't be running at >100c. When VRM fails, it usually fails without any warnings. Massive majority just installs the card and doesn't install Presicion X let alone use custom fan curve. I don't use Presicion X, I use Afterburner and it doesn't have any preset fancurves. Besides, the aggressive one in Presicion X is loud as hell judging by screenshots that people post.
 
The worse the case cooling, the bigger the problem. Especially if you do not do anything to the card. If everything was completely fine EVGA wouldn't kindly ship those thermal pads and update BIOSes.

If the VRM runs at 100 degrees  it will for sure have an effect on the GPU temps. If something gets really hot it will affect the surrounding areas = GPU!  About your argument about if VRM fails due to overheat. No (EVGA) card have failed due to overheated VRM. If you have some info about cards bricked, please post it. 

-_- No, VRM is far enough from GPU so that it doesn't necessarily show in GPU thermal sensor at all. VRAM is much closer and it does affect some of those chips though for sure. I never said that these cards have failed because of high VRM temperature. It is highly possible that it will happen in the future if the VRM keeps running at ridiculously high temps and the case cooling is weak. It might not fail now or even in near future. After one or two years though...
 
It has been shown in tests over and over again that VRM cooling is poor in these cards.

rsabatino
I got a 1080 founders edition. is there no bios change for me ? 

No. It's not needed because it has the stock cooler.

Anarion

emsir

Anarion

emsir

delicieuxz

emsir

Anarion

clone
I must be one of the lucky ones as my 1070FTW runs very cool but ive still asked for the thermal pads.

GPU might run very cool but the VRM are could be well past the boiling point.

This is rubbish, you have no documentation to prove your statement. Don't post something you don't have a clue about.

What do you find to be inaccurate about their statement? The temperature sensor on graphics cards is for the GPU chip, and not the VRM or VRAM. Whatever temperature is read from the GPU chip does not reflect the thermal conditions of the VRAM and VRM.
 
I hope that a cool GPU means also cool VRM and VRAM, but without thermal pads to transfer the heat from them, there seems to be significant risk to those components.

I've had my GTX 1070 FTW since September and played a lot of demanding games. No issues at all. So significant risk to components are not true. There haven't been 1 card with damaged VRM or VRAM. Nothing! And I think a lot of people with EVGA 10 series card are using agressive fan curve in Presicion X, and by that the card has no heat issue at all.  

LOL. That doesn't mean your VRM are wouldn't be running at >100c. When VRM fails, it usually fails without any warnings. Massive majority just installs the card and doesn't install Presicion X let alone use custom fan curve. I don't use Presicion X, I use Afterburner and it doesn't have any preset fancurves. Besides, the aggressive one in Presicion X is loud as hell judging by screenshots that people post.
 
The worse the case cooling, the bigger the problem. Especially if you do not do anything to the card. If everything was completely fine EVGA wouldn't kindly ship those thermal pads and update BIOSes.

If the VRM runs at 100 degrees  it will for sure have an effect on the GPU temps. If something gets really hot it will affect the surrounding areas = GPU!  About your argument about if VRM fails due to overheat. No (EVGA) card have failed due to overheated VRM. If you have some info about cards bricked, please post it. 

-_- No, VRM is far enough from GPU so that it doesn't necessarily show in GPU thermal sensor at all. VRAM is much closer and it does affect some of those chips though for sure. I never said that these cards have failed because of high VRM temperature. It is highly possible that it will happen in the future if the VRM keeps running at ridiculously high temps and the case cooling is weak. It might not fail now or even in near future. After one or two years though...
 
It has been shown in tests over and over again that VRM cooling is poor in these cards.

Well, if you have these tests of the VRM cooling, please post them here, so we all can see. Links would be fine. And where do have your information from that " It is highly possible that it will happen in the future". Are you sighted or have a magic ball you can look in to? And EVGA is not responsible for poor case cooling.

emsir
 
Well, if you have these tests of the VRM cooling, please post them here, so we all can see. Links would be fine. And where do have your information from that " It is highly possible that it will happen in the future". Are you sighted or have a magic ball you can look in to? And EVGA is not responsible for poor case cooling.

Either you didn't take the time to read the thread, where there is some of the links you're asking for (just some of it i would add) or you just don't want to use something as simple as google/bing/duckduck/or whatever to check by yourself.
Let me help you a little with an undeniable load of fundamental facts, you'll find incredible facts such as thermal resistance, junction temperatures, reliability, Thermodynamic, Physic laws, etc,  enough to "blow" or "Overheat" one's brain : simple as clicking here.
 
I agree with you about the case cooling, but EVGA is responsible with the card cooling, which is supposed to take the average environment during the design (best/worst).

GFAFS

emsir
 
Well, if you have these tests of the VRM cooling, please post them here, so we all can see. Links would be fine. And where do have your information from that " It is highly possible that it will happen in the future". Are you sighted or have a magic ball you can look in to? And EVGA is not responsible for poor case cooling.

Either you didn't take the time to read the thread, where there is some of the links you're asking for (just some of it i would add) or you just don't want to use something as simple as google/bing/duckduck/or whatever to check by yourself.
Let me help you a little with an undeniable load of fundamental facts, you'll find incredible facts such as thermal resistance, junction temperatures, reliability, Thermodynamic, Physic laws, etc,  enough to "blow" or "Overheat" one's brain : simple as clicking here.
 
I agree with you about the case cooling, but EVGA is responsible with the card cooling, which is supposed to take the average environment during the design (best/worst).

You got nothing to show. It's no way relevant to this issue what you are posting. If you don't have anything specific to EVGA cards  ( GTX 10 series), your argument is rubbish.  And you are completely wrong. EVGA is responsible for their products, NOT any average environment. EVGA has no influence where you install the card.  I don't see any threads with specific VRM issues and proof that VRM is overheated. If you took the time and read what EVGA post about this issue, you would know this:
 
"To resolve this, EVGA will be offering a VBIOS update, which adjusts the fan-speed curve to ensure sufficient cooling of all components across all operating temperatures.  This VBIOS will be released in the next few days and users can download it and update their cards directly.   This update resolves the potential thermal issues that have been reported, and ensures the card maintains safe operating temperatures.
 
For those users who want additional cooling beyond the VBIOS update, EVGA has optional thermal pads available.  This update is not required, however; EVGA will make it available free of charge to any customer who is interested. To request the thermal pad kit, please visit www.evga.com/thermalmod.
 
There is NO way EVGA says anything about overheated VRM. It' something some "experts" have posted and claimed. There is NO evidence to back this. That's why I want you to post these links that shows VRM is overheated. But you got nothing. 
 
 
 
 

TFG! You'll have your full refund and we'll be rid of you for good! If you had proof, you would show it. Otherwise, paranoia and BS!

DeathAngel74
TFG! You'll have your full refund and we'll be rid of you for good! If you had proof, you would show it. Otherwise, paranoia and BS!

TFG ?   Total Fishing Gear 

grendelrt
So now they are shipping VRAM pads as well, I am assuming this is only for if you have a gap? Or are they recommending to change out the vram pads regardless?

Answer at bottom, first some observations based on my experience of legal matters.
 
As far as I know EVGA are not recommending anything. They stick by their product and will keep their warranty agreements. Rather, they are offering all these "fixes," as purely a customers individual decision. It' understandable, EVGA can't put it stronger than that: Below is probably more accurate, but it won't happen. Next a different (imaginary ) legal/customer statement from EVGA):
 
      We have identified heating issues pertaining the to VRM/ fets on some cards. This can also affects the VRAM  chips (7) as they are close enough for heat transfer to occur. Further; in some cases the thermal tape on the VRAMS doesn't make full/or any contact with the plate above them. 
We assure you the GPU chip itself is very well cooled and we have not received or heard of any complaints regarding the GPU chip itself. The GPU is in direct heat transferring contact, with the whole finned cooling block and attached fans.
 The heat of the GPU chip is not a reliable way to check on VRAM, and VRM component temperatures. They have no thermal sensors, the GPU does have a thermal sensor.
In light of the above we deem the installation of additional cooling measures, and a BIOS flash necessary for full cooling of all components, and even just piece of mind. This is not to say it affects all cards, but we have numerous reports, and professional tests performed by several online IT companies and groups that do in fact show on some cards overheating of both the Fets and VRAM without overclock, running on original fan curve in the original BIOS. Some of these claims are dubious, we discount any that use only Furmark. But, a few are systematically  conducted, which leads us to this decision. Below:
   In light of the above we recommend customers to apply all or some of the free fixes. Those who do not wish to will still have full warranty. In addition, customers who do perform the fixes, and suffer accidental damage, will not have your warranty affected in anyway. We take fully responsibility for accidental mistakes related to the recommended procedures. In Summary:
1.  It is not a necessary requirement to do any of the safety cooling measures we are offering to all customers with affected skus. Warranty is not affected. We offer RMA for all customers.
2.  We recommend the installation of extra pads, in addition to the new BIOS. Rest assured, even if accidental damage occurs while performing these steps, your warranty is still completely valid.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
So in legalize (which I have to deal with as a part of my job) the above would open EVGA to all sorts of legal issues. It's mostly subtle so those without legal experience may wonder what's different. Please believe me, a statement like my fake example, in legal liability terms are very different in this world of litigation, and "legal crap".  I won't list them all! The above is pretty much saying there is a flaw in at least some cards, and it could be easily argued and almost certainly won by any lawyer. Believe me, lawyers would have a field day with this. (In the USA MacDonalds was sued successfully due to a lady spilling her coffee on her leg at a drive in. She had minor burns. She claimed McD are responsible for selling their coffee too hot. Laughable right!?! - But she won. McD lost the case!! You probably already know that story, but it is no longer unusual, particularly in the USA.) EVGA doesn't have the resources as a huge company like Samsung (Galaxy 7 issue.) Anyway the Galaxy 7 case is clearer so it's not a good example, but get the idea? I think they are doing the best they can, and EVGA are not doing anything illegal that can be sited with their statement. My imaginary one would!!
 
OK, sorry, just had to write that as it's clear a few posters don't realize the potential danger to EVGA of this whole sorry episode - hence a lack of admitting or even hinting at a flaw in their product. They simply cannot, for their reasons do that, ever, with things as they stand now.
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Finally, lol, grendelrt, to answer your question (summary below). Anyone can get the pads. If you look carefully and determine there is no gap between the original thermal pads and VRAM, my best guess is that they would be within acceptable limits. (This is opinion - I have nothing to back it up). It doesn't solve the VRM/fets issue though. Personally, I would do all three mods and BIOS for piece of mind.
 
In fact I have. I was disappointed, and shocked when I opened my card and found that the thermal pads on 7 VRAM chips were too thin. They were touching nothing. My case clearly needed fixing. Simply, the VRAMS were not getting thermal pad contact cooling at all!!
 
Final summary for grendelrts question: (At last, and it's my opinion only.)
A: If me, I would do it regardless. That's the best solution in my opinion.
B: If you see or even suspect any gap do it!
C: If you verify contact is proper and complete - it's your choice. I don't know well enough to speak with certainty in such a case.
 
Note: I expect different opinion, it's a forum after all!  But please read very carefully, or twice even, before typing. It's easy to mis-understand what I'm saying as it's not so simple, in particular, the legal bit. (which I simplified as much as possible.) I have dealt with different, but somewhat similar legal cases many times in my career. I am 48 year old man. So again, please read fully, understand fully before writing a response.
 
Sorry for the long post guys. I hope at least a few find it interesting or useful. FINISHED
=========================================================

Evga 1080 ACX 3.0 SC Back-plate removed!  Adjusted fann speed in msi afterburner.
 
Fann speed 30%=700rpm idle. And upp to 50% 1800rpm max fann speed at 70.c
 
Max temp with case laying down 70.c  with 50% fann speed
Max temp with case standing  58.c  with 50% fann speed