What's the issue with EVGA's ACX 3.0 coolers?

[benbaked]

I read on the Anandtech comments where people were griping about EVGA's ACX 3.0 coolers and how they were marketing-designed rather than engineering-designed and as a result causing VRMs to explode. What is all of that about? I have a EVGA 1060 FTW+ card that I purchased last November with the ACX 3.0 cooler, it's registered so I have the warranty on it. I'm hoping there's no cause for worry of it dying prematurely because of a poorly designed cooling solution.

 

[freakdiablo]

This is the first I've heard of it but I would really be surprised if the heatsinks were truly that bad. Yeah, flashy sells, but there's bound to be a hefty amount of QA that goes into these things. If it seams like a giant conspiracy, it usually isn't. If VRMs were going off like M80s due to inadequate cooling it'd go back for a fix.

 

[JrClocker]

It's all hoopla!

 

[cdawall]

If you run furmark for multiple hours on a 1070/1080 the VRM's do not get enough airflow to cool properly due to the cold running nature of the cards. EVGA offers both a thermal pad and an updated BIOS that completely fixes all of the issue related to it.

 

[TheCheat]

And here is the link for what cdawall is talking about:

http://www.evga.com/thermalmod/

 

[benbaked]

Thanks guys, my part number 06G-P4-6368-KR does not appear to be affected so I guess no problem, no worries.

 

[Kenrou]

It's all hoopla!

If you run furmark for multiple hours on a 1070/1080 the VRM's do not get enough airflow to cool properly due to the cold running nature of the cards. EVGA offers both a thermal pad and an updated BIOS that completely fixes all of the issue related to it.

Sadly it's not hoopla, this is the original thread we had running a few months back : http://www.overclockers.com/forums/showthread.php/776478-EVGA-GTX-1060-1070-1080-overheating-issues

"TLDR : in extreme circumstances like low air-flow and/or extended periods of heavy usage the VRM's might go above the temperature safety limit and cause severe damage to the GPU and possibly to the equipment around it. This only happens with EVGA ACX versions and the company has already provided VRM thermal pads for DIY install, a new BIOS to increase fan curve and RMA procedures."

 

[JrClocker]

No - it's hoolpa.

EVGA has stated and provided data that even though the VRM temp could be high in some cases, it is well within the design parameters. High temp does not mean that VRMs will explode or break, as long as the components are below their design thresholds.

EVGA issues the thermal pads as a free upgrade for anybody (I.e. those who were not comfortable with this statement).

The firmware upgrade does nothing more than provide a more aggressive fan curve on "auto". If you do your own fan curve this won't affect you.

 

[BuRgLaR]

If you're that worried about overheating just do what I do.


Whenever running GPU intensive stuff like gaming I just crank my fans to 100%

To quote Ron Popiel and the Showtime Rotisserie & BBQ infomercial SET IT AND FORGET IT!

 

[Kenrou]

EVGA provided data, the other websites using the damned things provided more data. Let's not start this again pls, just the whole idea of them doing so much if there's no problem is evidence enough of foul play, either a bad batch or an honest mistake of "forgetting" the VRM heatsinks comes down to the same, they made a mistake and owned it but downplayed it. Props for owning it, boo for downplaying it.

Again : Props for EVGA for owning the mistake, boo for downplaying it. If you have one of the earlier versions without the heatsinks and have a crappy case and/or fans you might want to get your free upgrade before starting heavy gaming sessions or bitcoin mining

 

[Nebulous]

Yeah I remember something about certain 1070/1080's VRM section overheating thus killing the cards. I believe it was caused by VGA bios' voltage set from manufacturer being too high and fans not ramping up high enough. Also was low quality VRM's being used. People were using MSI's Kombuster to bench up their cards only to find the card's VRM section could not handle the heatload even tho there were designed for something like 150c. Of course Kombuster's proggy exceeded that and pop goes the weasel.

EVGA corrected the issues with VGA bios updates and those cards that were affected were replaced.

Nothing to worry about now.