EVGA issuing thermal pads and VBIOS update for GTX 1080/1070 ACX 3.0 PWM and Memory

I know this is a stupid question but I am worried.

Could a GTX 1070 catch fire and burn down my house ?

 

I witnessed mine die live through the case window. Ever see a 4th of July sparkler? Kinda like that for a quick second or two. Unless you store volatile liquids or have a crap load of cat hair in your case...no.

 

Can you all please visit this thread I made:


http://forums.guru3d.com/showthread.php?t=410914


I need some very honest opinions please.

 

Those are probably the f'in sloppiest gpu cooler installs I've ever seen.

Come on, at least make the parts touch!

Disgraceful.

 

I got my pads today, but my RMA replacement should be here tomorrow. A friend of mine reported that his replacement card was in worse shape than his original card and he's not happy. I'll report back when I get the replacement. If the replacement is bad I'll send it back to EVGA and install the thermal pads on the card I bought.

 

Got a re-certified card as a replacement. I'm just going to send it back and install the Thermal Pads on my own card. Kind bummed with EVGA this time around and wishing I got a Strix.

 

Got my brand new GTX 1070 in a new manufacturer box Tuesday! Inspected it and looked like it had the fixes!

Sent back my old GTX 1070 in original manufacturer box in perfect shape.

 

I think I would have no problem buying from Evga in future. They've burned themselves badly this time around, so they must have learned a huge costly lesson: do not skimp on cooling design in any way shape or form. I was about to pull the trigger on a 1070 SC but was lucky to have read Hilberts review of the card (where he spotted the hot VRMs) before this issue exploded into public awareness, and that was what led me to avoid it.

I think all manufacturers, not just Evga, will now go the extra mile to ensure proper cooling of components. Since thermal imaging is becoming a more common thing in GPU reviews (I believe G3D were among first to use it).

 

I've installed all the thermal pads EVGA sent. The VRAM, the mid plate and the backplate ones. GPU temp is down with Thermal Grizzly on the die. I unfortunately do not have a infrared gun to measure the VRAM areas.

 

I got the thermal pads and bag of paste mod from EVGA last week but I've decided I am not going to fit them as I don't feel confident enough to dismantle a £600 graphics card, which is actually my second having already RMA'd the original due to the black screen/100% fan issue.

I know I'm covered and EVGA will replace the card if I damage it but I'd rather just use it and if it develops a fault again then RMA it since I have 2.5 years left on the 3 year warranty. It's highly unlikely that I will still own this GPU in a year's time never mind 2.5!

 

Get some confidence, because such a procedure is beyond easy.

 

Eh we shouldn't even have to do it. I took mine back to Microcenter and they took it back without asking questions, four months after purchase date. They're aware of what's going on and they fixed my issue unlike EVGA who gave me the run around. Microcenter me a brand new Strix OC. I wish EVGA luck in the future, but I won't blindly buy from them again.

 

Exactly. We really should not have to do this and since EVGA claim the thermal pad mod is OPTIONAL and just there for peace of mind then I am happy to leave it. EVGA will fit them if you want but to do so means having to send the card back and being without it for up to two weeks (which at this time of year could be three or even four!).

What annoys me about this is that I have a replacement GTX 1080 FTW from EVGA due to the black screen/100% fan issue on my previous card and rather than fit the pads before they sent the new one out they just left it as is for one of their customers to do. I even mentioned the thermal issue to support during the RMA but, nope, I was told that it wasn't needed and that the issue I had was an unrelated fault.

 

Then your VRM still isn't properly cooled. The thermal pad that goes on the VRM doesn't touch the heatsink or fins. Just look at a tear down of a Strix vs. a FTW. Made me mad enough to return the card. No one said it was difficult either, it's just a piss poor bandaid to a flawed cooler design.

 

I am confused. Unless you have a waterblock on the VRM, no manufacturer offers a big heatsink on the VRM as far as I knew. (Could be wrong tho) My pad touches the backplate which acts like a heatsink in a way as the plate now gets very hot. I even opened it up and flipped the pad to make sure it touches, and obviously, I saw indentations on the pads which proves they do touch. Maybe the pads could have been a few mms thicker but any more and they would just not transfer the heat properly IMO. 27C kind of proves for that specific card his application of the TIM and pads is working because most cards dont idle temp less than 30-32C...

 

27c is the temp of the GPU die, not the VRM. Need an IR thermometer to measure them.

 

And do you know where the VRMs are? Trust me, as I have the card and thoroughly tested various OCs with it, having the pads on and applying thermal grizzly, lowered temps and gave me unattainable OCs I had hit a wall at before, not mention, it also gave me a higher gaming OC with lower temps and stability.

Also, I had the strix,OC edition too. It couldn't even overclock near the same speed as the EVGA card does even before the pads and new bios update. Not by almost 100 mhz difference.

I think people are forgetting the warranty EVGA has on the card. 3 years is plenty of time as most people buying that card would upgrade by 2-3 years anyways. If it fails before then, I am sure a brand new card would be sent to you anyways. So the risk is way worth the reward to me.

 

Ouch.. even with thermal pads a 1080 blew up. Wonder if they're using sub-par resistors or other low grade components. Or the pads not making sufficient contact with the heat plate.

http://forums.evga.com/1080-FTW-exploded-now-dead-and-took-other-parts-with-it-m2583259.aspx

 

Well... My card is just reference design with custom cooler so I'm not worried.

 

Hmmm.... All of this happened on Z170... Thats not an extremely powerful rig but a good gaming build... I wonder if his motherboard had anything to do with it?