EVGA Overheating VRM Issue with GeForce ACX Coolers

Subject: Graphics Cards | October 30, 2016 - 03:09 AM |
Tagged: fail, evga

About a week ago, EVGA acknowledged an issue with their brand of custom-cooled GTX 1070 and GTX 1080 FTW cards. This came after the German branch of Tom's Hardware measured, back on October 6th, very high temperatures on the voltage regulator modules (VRMs), which was caused by these components not being able to adequately remove heat. To remedy the situation, EVGA offers cooling pads for all affected customers, which these customers could install under backplate and under the heatsink fins.

View Full Size

Image Credit: EVGA

Over the last day or so, users have been reporting that their cards are breaking, and even allegedly catching fire. According to GamersNexus and their source, Buildzoid of Actually Hardcore Overclocking, VRMs, if they fail, will just burn out without warning. The user in question claims that they were just playing Shadow Warrior 2 when their computer just shut down, with a sparkle and magic smoke. Taking the card out, they noticed a scorch mark on the PCB, right in the middle of the VRMs.

Regardless of how gloriously pyrotechnic this issue became, the consensus is that the thermal pads will still fix the issue. If you're not comfortable adding them yourself, then you should contact EVGA support.

Source: GamersNexus

Video News

October 30, 2016 | 03:41 AM - Posted by Anonymous (not verified)

Evga is a cheapskate and doesnt deserve the premium they ask for.

October 30, 2016 | 11:20 AM - Posted by Anonymous (not verified)

They were the ones that had coil whine issues on a previous generation GeForce (780, I think) because they didn't use a dab of epoxy on them like everyone else does. That is cheaping out.

October 30, 2016 | 03:43 AM - Posted by Anonymous (not verified)

Gamers Nexus is an absolute fucking joke of a dude/site, ROFL and that retard who is "his source", overcooked the card using custom build.

Fuck you for even allowing this shit post Ryan.

Stop sucking.

October 30, 2016 | 04:23 AM - Posted by Anonymous (not verified)

Maybe learn to read the entire post before posting a useless comment. Gamers Nexus isn't the original source, Tom's Hardware is. And the fact that EVGA has already responded by offering add in parts proves there is a legit issue. Yes the guy who says his card went up in actual smoke was likely running a custom setup, but that doesn't dismiss the fact there is an actual problem.

I've been using EVGA exclusively since I bought an 8800GT nearly a decade ago. But the company really has gone to hell, so this issue doesn't surpirse me in the slightest.

The one thing they've had going for them all this time is the fact that they sell their cards directly and offer their Step Up program to go along with that. But they shit the bed worse than any other AiB with their 1070 pricing. Take a quick look at it, it's a complete mess than makes zero sense. They have a special "Black Edition" of their 1070 SC listing for less than their basic 1070 ACX 3.0 with no factory over clock and no special color scheme.

The prices for their 1070 line are all over the map, and every single one, at least last time I checked, had a higher premium over MSRP than their corisponding 1080 cards. Which makes zero sense since they are using the exact same custom parts.

Heatsinks, fans, shrouds, every single part supplied by EVGA is exactly the same between the 1070 and 1080 cards, yet EVGA is charging more for those parts on the 1070s than they are on the 1080s.

This latest issue seals the deal, I'll never shop with them again.

October 30, 2016 | 10:23 AM - Posted by Anonymous (not verified)

Gaming GITs and their love of a brand, any brand! Really, what crawled up your nether region! You need some thermal pads and some water cooling! So go take a long walk of a short pier! And be sure to tie those concrete blocks on securely to keep yourself properly submerged!

October 30, 2016 | 05:27 PM - Posted by RT1 (not verified)

GN is easily one of the most technical of PC tech/gamer websites as those guys have a background in hardware. If you don't understand it, or cannot read well enough to understand that they are piggybacking off the absurd Tom's Hardware tests using a stress test that does not come close to simulating in-game temps, maybe don't call them a fucking joke.

October 30, 2016 | 07:55 PM - Posted by WhiteZero

GN does better hardware reporting/reviews than 99% of sites out there today. Nice shitposting though.

November 1, 2016 | 04:20 AM - Posted by Irishgamer01

Yeah, calling GN a joke just make u the laughing stock.
GN is one of the TOP sites which makes you a sad clown,
who should be pitied.

I assume you, like most apple fans love paying over the odds for decidedly average hardware.

(I do own an EVGA myself but I bought it on ebay at a discount).

November 1, 2016 | 09:18 AM - Posted by Anonymous (not verified)

wow......settle down?

November 2, 2016 | 10:13 AM - Posted by Anonymous (not verified)

EVGA themselves admit that there is a problem and are fixing it.

So yeah. It's a problem. An actual problem.

Grow up. Seriously.

October 30, 2016 | 03:46 AM - Posted by DAKON (not verified)

EVGA has gone the route of going cheap with the cards and attempting to be first in the line with custom cards. They have about no QC, especially when looking at this. And why is there no recall?

October 30, 2016 | 04:10 AM - Posted by Anonymous (not verified)

I've learned to stay away from EVGA, always had some kind of technical issue with their products in the past

October 30, 2016 | 04:20 AM - Posted by JohnGR

It's funny how a $400(1070) - $800(1080) card can end up in smokes because the manufacturer decided to save a few cents or something. Considering that you can buy a thermal pad 10x10 centimeters on eBay for about a dollar, the cost to the manufacturer is a joke. Not to mention that you can advertise those pads under the backplate and in fact sell a few more cards, because of those pads, thanks also to reviews measuring and showing temperatures.

Maybe we will see a new trend after this, with people adding thermal pads in cards(Nvidia or AMD) with backplates, if they don't have them already.

November 2, 2016 | 09:02 AM - Posted by razor512

They likely stopped doing proper thermal testing.

October 30, 2016 | 07:53 AM - Posted by Anonymous (not verified)

Time to investigate this like you did AMD when they had problems with the 480 pulling more power...knowing how biased you all are you won't want to piss off nvidia you shills.

October 30, 2016 | 10:30 AM - Posted by Anonymous (not verified)

Another gaming GIT that thinks a brand, any brand, is his mother. See the other post and do the same! Damn gaming GITs those brands are not your mothers! Your mothers work hard at the bunny club to bring home the dollars, one at a time, to keep you in that basement to hide the shame!

October 30, 2016 | 01:57 PM - Posted by Anonymous (not verified)

Awww did not shitvidia catch on fire so you start with the mom jokes?

They claim not to be biased this is a good time to prove lets see some tests done,like with AMDs 480s power draw issue.

October 30, 2016 | 06:26 PM - Posted by Anonymous (not verified)

AMD and NVIDIA both have their share of gaming GITs, and gaming has a GIT problem in general! Your(Red Team GITs) and the Green Team's GITs with your obsessive brand loyalty that leads to so much Butthurt!

Then there are the Morons talking about the new MacBook Pros with the Polaris 11 based PRO SKUs! Do you think they will be good for gaming when Apple is going for the certified Pro Graphics DRIVERS for Graphics(Non Gaming usage). The damn gamers do not even know the difference between the crappy gaming drivers that are tuned for quick and dirty and NOT very accurate gaming graphics and the certified Professional Graphics Drivers that are not tuned for flinging crappy gaming graphics at High FPS but are instead tuned for accuracy and high quality images.

Those MacBook laptops are not for gaming and never will be as Apple makes the drivers with the Help of AMD, and AMD is making the Polaris 11 based PRO SKUs drivers to its customer’s(Apple) specifications. I personally like that Apple went with AMD as AMD has more SP FP resources relative to the very little SP FP resources that Nvidia overcharges for. But still the gaming GITs are only interested in gaming so they can all F-Off and forget about gaming on the MacBook Pros because the drivers are not for gaming! And Apple is not going to white-list any other drivers but the Pro drivers that Apple/AMD made/certified to work with the professional graphics software that mostly have very little to do with gaming graphics.

That said good for Apple for using AMD’s Polaris, Bad for Apple with their F-ing thin and light sycophant obsession and the gimping of the new MacBooks to only 16GB of LPDDR3 RAM. MacBook Pros need 32GB+ RAM memory options on a fatter from factor! F-CK the battery life everybody that I see working with MacBooks at the coffee house has their MacBooks plugged in. Apple needs some really Pro Laptop SKUs with beefier Polaris 10 based Pro GPUs(Pro Drivers), or an Apple Branded and hardware white-listed external TB3 PCI dock for some really powerful Radeon Pro(Formally branded FirePro) Professional Graphics usage on the go!

A laptop with 32GB+ DDR4 RAM memory, PCIe 4.0(for the new SSDs/etc.), TB3 controller chips with link aggregation so more than 40 Gbs of bandwidth is available for any external TB3 PCIe external GPU enclosure using two TB3 cables, and so on!


"Git is a mild[1] pejorative with origins in British English for an unpleasant, silly, incompetent, stupid, annoying, senile, elderly or childish person.[2] It is usually an insult, more severe than twit or idiot but less severe than wanker, arsehole or twat.[1][3][4]

The word git first appeared in print in 1946, but is undoubtedly older.[citation needed] It is originally an alteration of the word get, dating back to the 14th century.[citation needed] A shortening of beget,[5] get insinuates that the recipient is someone's misbegotten offspring and therefore a bastard.[6] In parts of northern England, Northern Ireland and Scotland get is still used in preference to git"

October 30, 2016 | 11:57 PM - Posted by arbiter

wow love people that are complete tool's. Issue AMD had was a problem on pretty much every card no matter WHO made it. This is a problem with 1 AIB maker.

need to ban "Anonymous" posters, 99.999% of time nothing but troll's.

October 31, 2016 | 05:24 AM - Posted by nonymous (not verified)

You are the biggest nvidia troll in here. You should be banned.

October 31, 2016 | 04:55 PM - Posted by Scott Michaud

Ideally no-one would be banned, and we all learn to treat each other respectfully. Unfortunately, a lot of trolling is caused by escalation when people don't believe they will be heard otherwise, which banning exacerbates, not remedies.

Even though we leave our comments open, and many of us do read them, it's difficult because there's often nothing we can do to appropriately respond to them. Humans are really good at spotting patterns, but really bad at identifying them; they usually assume intent, which is when they go into "conspiracy theory" territory. You're then viewed as "a bad person" and it snowballs.

October 31, 2016 | 05:30 PM - Posted by arbiter

Its really sad though scott how they are going about this issue on eVGA cards which seems to limited to FTW cards and turning it in to its ALL of the nvidia based cards made have this issue even ones from MSI, ASUS, gigabyte, etc. This isn't the issue RX 480 had which was all cards since it was a Bios issue of drawing to much power from the board.

October 31, 2016 | 05:36 PM - Posted by Anonymous (not verified)

If this is sad then you must think your post in forums are pathetic.

Double standard / Hypocrite much ?

October 31, 2016 | 05:44 PM - Posted by Scott Michaud

Be nice, folks.

October 31, 2016 | 05:44 PM - Posted by Scott Michaud

Yeah, that's not really your problem, though. It only affects EVGA regardless of what people say.

October 31, 2016 | 06:50 PM - Posted by Anonymous (not verified)

I don't care what vendor the problem is on it's still an Nvidia product.now if it was an AMD gpu they would have been roasted (pun intended) with a three page report with all sorts of tests done. to show how serious the issue is but because its's Nvidia,that won't be done and it's no wonder some claim bias toward Nvidia.
Again because it's Nvidia the issue gets down played by Nvidia shill reviewers.

October 31, 2016 | 07:49 PM - Posted by Scott Michaud

Have you thought about how this sounds to reviewers who genuinely believe they're doing things unbiased (although really every human is biased in several different ways)? Even if you noticed a pattern, you're claiming that people are intending something. What if they aren't?

October 31, 2016 | 08:07 PM - Posted by Anonymous (not verified)

Untill theres a 2-3 pages report with alot of tests done to see what serious this issue is.i don't care how that sounds.

October 31, 2016 | 08:08 PM - Posted by Anonymous (not verified)


October 31, 2016 | 08:09 PM - Posted by Anonymous (not verified)

I don't thinks its only him. People in the community (check other forums) pretty much think PCPerspective is Nvidia BIAS. Even if Ryan himself thinks hes un-bias that many people noticing a pattern with him isn't by mistake.

October 31, 2016 | 09:18 PM - Posted by Scott Michaud

That's not what I'm saying.

Think about what will happen if you accuse someone of intending something they aren't. Not only are you removing your credibility for your argument, but you're also associating it with negativity. It takes a lot of professionalism (assuming they're even aware of it) to negate that negative bias that you inflict upon them.

October 31, 2016 | 10:12 PM - Posted by Anonymous (not verified)

I don't think credibility of those accusing is in question when there is a pattern that is there for all to see.

Bias reputation isn't something invented especially with a techsites across several forum communities. That type of bias reputation is earned.

October 31, 2016 | 11:21 PM - Posted by Scott Michaud

That doesn't work when the accusation is intent and it's attempting to convince the very person that you believe is intending something, when they aren't.

October 31, 2016 | 11:36 PM - Posted by Anonymous (not verified)

Actions speak louder than words...you can keep saying pcper isn't biased tho.what ever makes you feel good.

November 1, 2016 | 12:18 AM - Posted by Anonymous (not verified)

You know what doesn't work DENIAL

November 1, 2016 | 12:33 PM - Posted by Jeremy Hellstrom

IT'S ALL TRUE!  Even though I have never owned an EVGA product since the GTX 670 nor ever owned an ACX Cooler I am supporting them with all my heart in the knowledge that they love me like the son they never had!!!!!  If anyone here ever tries to post negative things about my one true love I delete them immediately!!!!!

You know why?  It's because the plural of ancedote is fact and showing bias means not agreeing with me and doing exactly what I say immediately!

November 1, 2016 | 01:41 PM - Posted by Anonymous (not verified)

You don't have to agree with these guys just check your reputation erg PCPerspective and Ryans on other tech sites forums across the world.

A bias reputation is hard earned.

November 1, 2016 | 05:29 AM - Posted by JohnGR

"That's not what I'm saying.

Think about what will happen if you accuse someone of intending something they aren't. Not only are you removing your credibility for your argument, but you're also associating it with negativity. It takes a lot of professionalism (assuming they're even aware of it) to negate that negative bias that you inflict upon them."


The only software on the planet that shows any kind of indications that Pascal cards have any kind of Async support is 3D Mark Time Spy. Many where making their own analysis showing that the program was not taking advantage of AMD's hardware but it was written specifically to show positive results on both Pascal and AMD cards. What PCPer did? Well it did reported the problem - after comments in the comments section, but before 3D Mark's announcement about it - and then did a pod cast with a representative of 3D mark.

I am pretty sure you haven't noticed something bad until here. Here is the funny part. I was looking at that podcast and in a few words I was looking Ryan and Allyn(I think it was also Allyn) with the 3D Mark representative laughing at those insisting that the benchmark was written specifically so to take advantage of Nvidia's Dynamic load balancing, but NOT AMD's hardware. The main argument if I remember correctly from Ryan "They are 3D Mark, they are making benchmarks for 20 years, they know what they are saying". Translation: "He is a politician for 20 years, trust him, he is saying the truth". Believe me that facepalm was painful.

November 1, 2016 | 01:41 PM - Posted by Scott Michaud

Yeah... I usually err on the side of free expression, but the replies to this post were nothing but empty hate, so I nuked them.

November 1, 2016 | 04:16 PM - Posted by JohnGR

No problem. :)

November 1, 2016 | 05:11 AM - Posted by JohnGR

"Have you thought about how this sounds to reviewers who genuinely believe they're doing things unbiased (although really every human is biased in several different ways)? Even if you noticed a pattern, you're claiming that people are intending something. What if they aren't?"


He is right, or to be more fair I believe the same thing and I was making some really loud noise in the comments in the past.

IN THE NOT SO DISTANT PAST(I will explain latter the caps lock) if this was Sapphire for example, there would have been much more articles out there attacking with their conclusions not just Sapphire, but AMD also. The problem with the Micron chips on Pascal cards? It would have been hell all over the place BEFORE the new firmware is issued. Not radio silence. I have seen this before. A problem, ANY problem with an AMD card, 3-5 pages analysis with AMD responding with a fix only a few days latter. Tech press was rushing to show the problem before AMD have the time to fix it. Nvidia problems? Months latter just an article to inform that the problem was finally addressed.

In the case of the RX 480 PCIe fiasco(it was a fiasco) I was expecting/asking/screaming at the press to check one of those GTX 950's without a 6pin PCIe connector that where seems to not being power limited. No site even bothered. All sites, this included, conveniently where testing a GTX 960(if I remember correctly) that DID HAD a PCIe power connector. Testing an 120W card with an extra six pin power connector. Convinient if not funny.

Now why the capitals? Because AMD is recovering. Zen could be a success, even if it is slower than latest Intel cpus(and it will be), Vega could be a success, Zen APUs and future Intel integrated GPUs could limit even more Nvidia's playing field, leaving them only the market over $200 in the beginning and $300 latter(except if they finaly decide to bring Tegra to the laptops/desktops). With AMD not only not going bankrupt, but recovering, the tech press seems to have change it's attitude towards that company. So the press had stopped attacking the company with so much delight. Yes, Nvidia and Intel do have better tech coverage and protection, but attacks to AMD have been much lower than before. Press is taking a step back because they don't want to be in a difficult position after 1-2 years, having to buy Zen and Vega samples because AMD is denying sending them.

October 31, 2016 | 08:15 PM - Posted by arbiter

Um if AMD had a problem that was limited to 1 maker that maker would take the assault, but any issue that has come from amd cards though has been ref cards which are identical over all cards, example 290(x) ref cooler dropping 20% performance or rx 480 drawing to much power from the board slow. So you claiming amd gets trashed while nvidia gets pass is wrong since mostly all the cases you can claim haven't been the same. nvidia took flakk over 970 issue. the memory issue was still party on the makers since some branched out using memory from diff maker though it still was on nvidia cause bios wasn't right to deal with it.

November 1, 2016 | 05:35 AM - Posted by JohnGR

The 970 was something that could not be covered up. And even then the press was trying it's best to convince people that the 970 was still the same good card they bought. Yes it was. But it was not the card advertised. And it is much easier to show this today, with 970 loosing much more in modern games compared to the 980. If the 970 had the specs that where advertised when it came out, it's performance would have been more consistent and closer to that of 980's.

November 1, 2016 | 04:42 AM - Posted by JohnGR

"October 31, 2016 | 05:30 PM - Posted by arbiter

Its really sad though scott how they are going about this issue on eVGA cards which seems to limited to FTW cards and turning it in to its ALL of the nvidia based cards made have this issue even ones from MSI, ASUS, gigabyte, etc. This isn't the issue RX 480 had which was all cards since it was a Bios issue of drawing to much power from the board."


EVGA is considered - if I am not mistaken - the top brand for Nvidia cards and of course EVGA is selling only Nvidia cards. So it's no surprise that for some when looking at EVGA it's the same as if looking at Nvidia. Think Sapphire and AMD.

October 30, 2016 | 08:14 AM - Posted by Mad_Dane

Shit like this only pass QA on purpose, there is no other explanation! I mean FFS making sure the heatsink touch the hot parts, how hard can that be to figure out!

October 30, 2016 | 08:35 AM - Posted by DrB (not verified)

Come on guys. It's coming sense. Polarity have to stay away from each other . You think trucking industry separate positive and negative for the lol of it? You think we use battery most opposite pole to charge battery or discharge equally . Come on
It's because of induction . Bottom line, positive connector shouldn't be close to negative connector and wiring should be done so negative and positive are away from each other. This happened because of induction.

October 30, 2016 | 08:40 AM - Posted by Mad_Dane

What does the positive and negative pole of a battery have to do with a heat sink not touching the all the hot components on the PCB?

October 30, 2016 | 11:32 PM - Posted by Tim Verry


October 30, 2016 | 10:20 AM - Posted by Anonymous (not verified)

I'm evga fan but both my 1080 went bad not even 90 days in then they tried to send me referb replacements I fought and got new ones. Then FTW black screens and now VRM catching fire. They really dropped the ball this time and making have second though using them again.

October 30, 2016 | 10:54 AM - Posted by Anonymous (not verified)

EVGA has really down hill :(

Although I used to be an EVGA fanboy, I always knew their cooling was sub-par because you can't mine with ACX cards without blowing your VRMs pretty quick.

October 30, 2016 | 11:25 AM - Posted by Lemonsquare

I would be careful here. My understanding is that these conditions are occurring only under extreme artificial benchmarking ie furmark, which is obviously not representative of playing games. But yes, EVGA should've just included the thermal pads from the get-go.

October 30, 2016 | 03:31 PM - Posted by Scott Michaud

The recent post on Reddit claims (with pictures) that they had a VRM blowout when playing Shadow Warriors 2.

Of course, it's possible that, when it became an issue, someone could have intentionally overclocked and ran Furmark / OCCT to try to make it burn out. It's best to assume that's not the case, though, especially since it shouldn't have been possible in the first place. Let's just assume someone had a bad day, and EVGA tech support will take care of them.

October 30, 2016 | 11:31 AM - Posted by Anonymous (not verified)

Those C__ts at EVGA! Just to save F-ing Two Bits they don't use any pads and now they are bleeding everywhere! WTF with FTW, and now the RMAs and the red ink!

October 30, 2016 | 02:34 PM - Posted by TooL (not verified)

I received my 1080 FTW in late July and have been playing BO3, BF4 and now BF1 with no issues.

No coil whine. Temps rarely, if ever, reach 70°C playing Ultra settings 2560x1440 @140Hz

I'd guess most of the posters here don't own one, but like to complain anyway.

That being said, I will order this cooling pad. Just in case.

October 30, 2016 | 03:47 PM - Posted by Anonymous (not verified)

nvidia cards don't have VRM temp sensors.

October 30, 2016 | 11:59 PM - Posted by arbiter

wccftech (yea i know) story tried to push this as it was a card wide range issue. Mostly all stories i seen are limited to FTW cards which i have the SC card. Had it so far for 4 month's no problem's.

October 30, 2016 | 05:35 PM - Posted by RT1 (not verified)

As noted, using Furmark to simulate in-game temps is a fool's errand and Tom's should know better. EVGA hot boxed their card before release and found no issues using the mid-plate as passive cooling. They made a mistake in not realizing that people would stress their cards beyond acceptable limits, and that there are many people with deep pockets who apparently don't understand the need to put direct air on the hottest component in your case - something the vast majority of cases easily allow you to do. That has become less of a thing with front mounted radiators as everyone needs water in their PCs now. That is going to hurt the ability to get significantly cooler air into your case and onto that hottest component. That was a rule of thumb born many years ago when cards were much hotter but it is still just common sense.

I have never seen my 1070 get over 65c. Not in the dead of summer in an 85f room because I am too cheap to pay the electric company. I also have two front fans less than 7 inches from the card, blowing on the card with a custom fan curve based on case temps.

EVGA's lesson? Don't count on people to be logical. That should be totally evident at this point in time.

October 30, 2016 | 07:03 PM - Posted by Cory Brown (not verified)

What was your VRM temps duing those high temps.

Love me some EVGA, a oversight in my opinion

October 31, 2016 | 12:02 AM - Posted by arbiter

Sadly people don't realize that software like Furmark push's every bit of card 100% which in real world use in mostly any game doesn't do that.

October 31, 2016 | 03:41 AM - Posted by Mad_Dane

Let me turn this upside down, if your car caught fire if you pushed it to 100% how would you feel? Then the automaker would proceed to tell you this doesn't happen in normal city and highway driving.

Any consumer product should be able to be handle to be pushed to the limit, no matter what software you throw at it! There is no gaming only disclaimer when you buy a GPU!

October 31, 2016 | 04:58 AM - Posted by arbiter

I have pushed my eVGA ACX card to 100% and then some many times over 4 months I owned it guess what NO fire.

Furmark still push's the card more then what more normal daily use pushes it.

October 31, 2016 | 05:10 AM - Posted by Mad_Dane

So what, any GPU should be able to handle furmark with out catching fire!

No software should be able to melt a GPU, then the hardware or firmware is designed wrong!

October 31, 2016 | 05:27 PM - Posted by arbiter

Its a story about 1 card from an AIB maker. Stop being an idiot and making it that its ALL gtx1080/1070's which seems to be what AMD fanboyz are trying to do.

October 31, 2016 | 10:42 AM - Posted by Anonymous (not verified)

I'm very happy for your fortunate.

Though, I still feel bad for the unfortunates.

October 31, 2016 | 09:08 AM - Posted by Cory Brown (not verified)

I'll take that bet, even wager a weeks pay on it. Take your automobile to the track and if it breaks there i bet the automaker will weasel out of the repair due to user abuse. Cheers and have a nice week without any pay! :-)

October 30, 2016 | 07:04 PM - Posted by CNote

My EVGA 970 SSC acx would get really hot if I didn't start precision x eventhough it's supposed to start when start up. Haven't noticed anything with my new EVGA 980 ti SSC acx

October 31, 2016 | 05:43 AM - Posted by arbiter

default fan curves on every card tend to be little lacking.

October 30, 2016 | 10:41 PM - Posted by Godrilla (not verified)

Evga used to be good, I actually realized that they were almost conceded in terms of number one Nvidia vendor for years especially with holding oc records kingpin line. They got to the point where they realized we can cheap out on parts and people will still purchase because of name and loyalty. After the whole forum went nuts over 2nd gen Maxwell 970 and 980 coil whine I went with gigabyte for the 980 g1, then zotac amp extreme 980ti, and now amp extreme 1070 till 1080ti.

Evga is no longer number one Nvidia vendor all vendors have their own premium versions, do your research, check out forums and newer reviews before purchasing.

October 30, 2016 | 10:45 PM - Posted by Godrilla (not verified)


Fyi before Maxwell 2.0 and only owned evga cards 8800 ultra, 285 ftw, 480s sc sli, 580 classified, and 690 classified .
Evga just lost a loyal customer years ago and now for good.

October 31, 2016 | 06:12 AM - Posted by khanmein

credit to GURU3D not tomshardware or wccftech or gamers nexus.


where's the micron review for GTX 1070???? thanks.

October 31, 2016 | 08:06 AM - Posted by JohnGR

In Guru3D's article it doesn't talk about an FTW card. The problem was spotted on an FTW model.

October 31, 2016 | 08:23 AM - Posted by khanmein

all the ACX cooler got issue. this time round EVGA got issue & stop covering them up & where's the micron vs samsung review? please don't over-shadow by the VRM over-heating cos this matter is pretty small.

whereas tech reviewer don't have guts to make a comparison. come find me a GTX 1070 with micron review at youtube by any well known reviewer like jayztwocents, linustechtips, hardwarecanucks, bitwit, paul h/w, tech of tomorrow, joker productions, tech city, dans tech, etc.

so far i only read 2~3 articles that reviewed micron.

October 31, 2016 | 12:30 PM - Posted by JohnGR

You mean the memory problem on the latest GTX cards? I think that was fixed with the newer firmwares. But it is funny to see things like these pass tech quality check. Micron chips don't getting enough voltage to be stable and VRMs getting hotter than what they should and going in smokes. You guess that things like that are the reasons engineers are getting payed for and yet they pass "quality" control. If we add one more incident with Nvidia removing a driver for being buggy, this was a busy month for Nvidia.

November 1, 2016 | 12:23 AM - Posted by khanmein

my question is reviewer received golden sample can consider acceptable but how come there's no micron review on the youtube?

GTX 1070/60/50/50Ti reviewed with samsung chip whereas we as consumers ended up with micron = crippled.

the new vbios may look like fixed but not. y so troublesome using micron? don't said that the production issue. y can't u blamed tech reviewer is bribed by them?

find me a video at youtube that make a neck to neck comparison e.g. like Asus Strix GTX 1070 OC (samsung) vs Asus Strix GTX 1070 OC (micron)

if u blamed board partners, then y now FE started to ship micron? the samsung production is stopped??

November 1, 2016 | 05:40 AM - Posted by JohnGR

If I was a reviewer I would be thinking that my golden sample came with Samsung chips because they are more overclockable.It wouldn't surprise me.

But if there is a pattern where companies send cards with Samsung chips, while using Micron in their retail parts, tech press should start investigating it. For every brand out there, no matter the GPU manufacturer.

To tell you the truth I haven't thought about that before.

November 1, 2016 | 10:08 AM - Posted by khanmein

this type of tech reviewer also can review?


summoned @Allyn Malventano to assist joker slunt.

if u asked him regarding micron issue. he said i don't know cos i bought GTX 1070/60 & RX 480 came with samsung expect GTX 1080 micron.

steve burke from gamers nexus said this is very common changing other vendor parts but c'mon if u guys insisted no different with a new vbios fixed everything then make a video.

this type of video can attract a lot viewer. like i said PCPER must do a proper honest legit review regarding this issue. the VRM over-heating is not a big deal cos the price is slightly lower compare to other big guns.

October 31, 2016 | 02:25 PM - Posted by Anonymous (not verified)

Burning down da house!

October 31, 2016 | 04:24 PM - Posted by Anonymous (not verified)

EVGA has excellent customer support though which has swayed me to be a customer.

Perhaps look at this as a TEST of how the company responds to an issue rather than condemning them for having an engineering design error.

They acknowledge the issues and created solutions in a relatively timely manner, and had your GPU died in the meantime the Warranty procedure is fairly painless.

*Like many companies however, you need to REGISTER your product within 30 days or you may limit your Warranty. In some cases it doesn't matter. (You can also EXTEND to 5-year for 5% or 10-year for 10% extra which is quite reasonable.)

**I decided to get the thermal pad kit from EVGA and apply it. Even if I don't overheat it may help reduce premature wear.

November 1, 2016 | 02:46 PM - Posted by maonayze

Personally I have never seen a post by any AMD Fanboy or otherwise that has said this is ALL 1070/1080 cards with this fault. As far as I am aware this is currently affecting EVGA 1070/1080 FTW cards. If it does affect any others I have seen nor heard reports of it....yet!

As for testing and doing videos then yes I see no reason why PCper shouldnt do some testing on those specific cards to check those VRMs for extremely high temps while gaming.

Just trying to get to the actual facts of this issue and trying to understand exactly what the issue actually is, before getting deeper into it.

As I understand it, the problem is the fact that the VRMs were not covered with thermal pads. Am I wrong in thinking that ALL gfx cards regardless of make/model are supposed to come with some form of heatsink for the memory and VRM chips and EVGA for whatever reason have left them out of an unspecified number of FTW cards?

On the odd occasion, I do get the feeling that PCper do come down on AMD and do more digging when a fault is discovered on the red teams side. It also doesnt help that this was first discovered on Oct 6th and the first we know about it here on PCper is Oct 30th. Again if this was an RX460/70/80 card I feel it would have been reported here on Oct 7th at the latest... It's just another piece of evidence that could be used by someone claiming that PCper is bias to one company over the other.

So lets show a bit of balance and do the same level of investigative journalism on both sides of the fence. After all, I think this kind of issue regardless of scope is more damaging than the extra power being pulled from the Pcie port of RX480s.

[On a sidenote to this, Steve from Gamers Nexus did a test on the RX480 on an old motherboard in the first part of a supposed two part feature about the power draw.....the second part of the video never appeared because nothing happened....this was covered by an addendum at the end of an unrelated Gamers Nexus Video and stated that nothing happened and the RX480 did not explode or overheat, thus brushing it under the carpet when the results didnt go the way they expected/wanted for maximum clicks or impact - just another Freudian slip by a gaming channel that no-one seems to have noticed.(Bias....possibly,as it was evidence that nothing would happen to those cards from the risk of overdrawing power from the Pcie Slot. In reality.... who knows.)]

Sorry for the long semi-rant but really only asking for a level playing field with the same amount of attention to all sides, that's all. :-)

November 2, 2016 | 12:09 PM - Posted by Anonymous (not verified)

Ya goog luck with that they will just cover this up to collect that check from nvidia.

November 2, 2016 | 06:04 PM - Posted by Jeremy Hellstrom

That could also show our bias towards AMD ... if you drop assumptions such as us doing this without contacting the companies first and then awaiting replies or even recalling what our actual conclusion to the power draw issue was.

Your timing estimate is also off, no way could a test be designed, performed and replicated overnight. 

We have heard from Jacob and I am trying to get a post up but this weird thing called a day job is interfering ... though I don't have small infants to care for or an entire office to move and furniture to build so I have that going for me.

November 2, 2016 | 07:10 PM - Posted by Anonymous (not verified)

Like the poster said above this was first discovered Oct 6th..so you're saying its taken a month to get a reply from evga?ya right nice try.if this was AMD Oct 7th we would've heard how bad the issue is and so on.

November 2, 2016 | 07:45 PM - Posted by Jeremy Hellstrom

Nope, pretty sure there is no mention of Jacob taking a month in that reply.  

Ignoring that and moving on, which day was it in June that we started talking about the voltage we were seeing being drawn?  I'll give you a hint, it was long before Ryan published on June 29th ... and this probably doesn't fit with your memory but what was the conclusion he drew?

As is usually the case in this market, there are some asking for a recall on the RX 480 or saying that AMD is going to be liable for some kind of class action lawsuit. I don’t agree with either sentiment. AMD is clearly pushing the envelope with the power delivery system on the Radeon RX 480 and we have shown that they are definitely drawing power outside PCI Express specifications even at stock clock speeds and power settings. I haven’t run into any stability concerns yet nor have I heard of any other reviewers indicating as much.

Weird eh?  but wait ... what happened on the 1st of July ... we couldn't have done the exact same thing to an NVIDIA card because of reports of power spikes on it now could we?  No, that's unpossible.

But don't worry, you are always right and everyone else is wrong if they disagree with you.

November 2, 2016 | 09:15 PM - Posted by Anonymous (not verified)

I really like watching you dismantle the trolls who like to accuse PCPer of being shills.

November 2, 2016 | 11:22 PM - Posted by Anonymous (not verified)

First of all he did not dismantle anyone.second its funny you choose to pick the 960 as an example...when he down played it as not being as serious...again Nvidia shills at their best.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

This question is for testing whether you are a human visitor and to prevent automated spam submissions.