EVGA responds to hot VRM area on GTX 10 series

[Kinda Bottlenecked]

1 hour ago, aisle9 said:

So no reason to reconsider my Gaming X purchase?

 

Nope. But if you'd like better temps, Palit Gamerock has a pretty cool one. 

[aisle9]

5 minutes ago, Pohernori said:

 

Nope. But if you'd like better temps, Palit Gamerock has a pretty cool one. 

I would love to get my hands on a Palit. Unfortunately, they don't sell in the US...although I believe PNY is owned by Palit?

[potoooooooo]

14 hours ago, dacheezypuffs said:

You sound like you're just jealous and have no knowledge of what "quality" is in a GPU. Just because they made a mistake doesn't mean they are the first, nor the last to do it and it doesn't make it trash.

I do know what quality GPUs are, and they typically don't catch fire even after the official remedies are applied

[zMeul]

11 hours ago, aisle9 said:

PNY is owned by Palit

PNY and Palit have nothing linking them

 

PNY is an american company mostly selling in Europe

 

Palit is chinese corp that owns quite a lot of brands, but not PNY

[aisle9]

14 minutes ago, zMeul said:

PNY and Palit have nothing linking them

 

PNY is an american company mostly selling in Europe

 

Palit is chinese corp that owns quite a lot of brands, but not PNY

Who am I thinking of that owns PNY? 

[zMeul]

5 hours ago, aisle9 said:

Who am I thinking of that owns PNY? 

"no one" owns PNY, PNY owns PNY

[Bleedingyamato]

16 hours ago, aisle9 said:

I would love to get my hands on a Palit. Unfortunately, they don't sell in the US...although I believe PNY is owned by Palit?

Is there something about Palit graphics cards that makes them so appealing to you?

 

[aisle9]

1 hour ago, Bleedingyamato said:

Is there something about Palit graphics cards that makes them so appealing to you?

 

Reputation combined with the novelty factor.

[zMeul]

GTX1080 FTW bit the dust with BIOS update applied: https://www.reddit.com/r/nvidia/comments/5e5723/evga_gtx_1080_ftw_exploded_vrm/

 

 

as the user tells, 

Quote

I had also done the bios update on the day it was released.

...

The 'optional' thermal pads arrived yesterday, the plan was to fit them today. So no, there were no pads fitted at the time. Since they stated that the card was fine to use without them with the bios update, I thought I would have been ok until they arrived. Clearly not the case unfortunately.

 

---

 

jeez H christ, I never had the chance to buy an EVGA card but hell if I'll ever do after this shit

[Bleedingyamato]

11 minutes ago, aisle9 said:

Reputation combined with the novelty factor.

What's novel about their graphics cards? 

[aisle9]

20 minutes ago, Bleedingyamato said:

What's novel about their graphics cards? 

Have you seen their coolers? Pure sex. 

[zMeul]

2 minutes ago, aisle9 said:

Have you seen their coolers? Pure sex. 

[Bleedingyamato]

6 minutes ago, aisle9 said:

Have you seen their coolers? Pure sex. 

I might've heard of the brand but that's it.  

[DrMikeNZ]

1 hour ago, zMeul said:

GTX1080 FTW bit the dust with BIOS update applied: https://www.reddit.com/r/nvidia/comments/5e5723/evga_gtx_1080_ftw_exploded_vrm/

 

 

as the user tells, 

 

---

 

jeez H christ, I never had the chance to buy an EVGA card but hell if I'll ever do after this shit

Thermal pads arrived last night and were promptly installed. Quite easy and fun to do.

 

I have noticed however that the fan profile (which was apparently supposed to be changed by the BIOS update) is still the same as it originally was when it isn't overridden by Precision OCX.

Perhaps the BIOS update didn't do what it was supposed to.

[Nowak]

This is becoming ridiculous.

[Rune]

2 hours ago, DrMikeNZ said:

Precision OCX.

That'd do it. Overwrites the fan profile stored if you have made any changes.

[dacheezypuffs]

14 hours ago, potoooooooo said:

I do know what quality GPUs are, and they typically don't catch fire even after the official remedies are applied

Yeah that's not catching fire. It's still cooler than an R9 290X.

[DrMikeNZ]

7 hours ago, Rune said:

That'd do it. Overwrites the fan profile stored if you have made any changes.

I had uninstalled Precision OCX when I checked the fan profile (informally by using HW monitor to check fan speed while benching the GPU), and after I re-installed it I immediately reconfigured it to my very aggressive fan profile (so fan speed starts at 60%).

I would be concerned if Precision OCX was causing the fan profile to not work correctly when it is not running and no longer installed.

[Mira Yurizaki]

So Gamer's Nexus tried to do their damnedest to get a card to blow up, but in the mean time, they tested using various configurations with Furmark (which is unrealistic) and Metro 2033.

The gist is:

• This is not a temperature problem. Gamer's Nexus' test at their worst case (given 40C or so of directed airflow) resulted in the PCB getting up to 120C but the VRMs stayed at about 110C and the card survived for at least 1000 seconds. This was also running Furmark. Given a normal use case, even without the thermal pads and BIOS, the VRMs were topping in the 90s. They also measured using thermal couples at key hot spots.
• The VRMs can tolerate temperatures up to 125C.
• The chokes (which is what one of the thermal pads covers on the heat sink side) can tolerate a ton more heat since it's just copper winding (of course, if the insulator material melts that's another story but at that point the card is probably toast anyway).
• They went into discussion on some myths, one of them being the thermal padding on the heat sink side makes things worse because it blocks airflow to the VRMs. The point they made was that "air channel" is so restricted anyway by components it doesn't matter, there's already a thermal pad over the VRMs, and adding the thermal pad improves the thermal conductivity significantly anyway.
• EVGA reported that their defect rate is 200 parts per million, or 0.02%. Basically, for every one million cards, 200 of them have showed up defective. This is within acceptable tolerance for the consumer world (i.e., the FTC isn't going to get on EVGA's ass)
  • The caveat is of course this is a number reported by EVGA
  • The other caveat is that this number may only be counted for if people actually submitted their cards through the RMA process, so it may be higher.
  • EVGA is suggesting the reason why you hear about it so much is because this is the way the internet works, and some people may have cards that aren't a problem, but there's something else wrong with it and because this is all over the news, they think the card is at fault.
• They had a conversation with someone (Bill Zoid? I dunno) mentioned that it's most likely the capacitors and it's likely EVGA received a bad batch.
  • The explanation Bill gave was that the capacitors get damaged somehow (overcurrent perhaps, which is likely when you're trying to overclock the card hard) and the capacitor starts to degrade. When a capacitor fails, it fails to one of two ways depending on the type: open or shorted. The type used in this case fails to a shorted state (and no, you cannot simply get a capacitor that fails to an open state because the characteristics change) and the capacitor blows. This is independent of temperatures
• Their conclusion is manufacturing defects or similar. They still gave knocks on EVGA for not having the thermal pads and BIOS already.

So there's the data, there's your information. Whether or not you want you still want to purchase or recommend a GeForce 10 EVGA card is up to you, but at least go in informed and not as a blind sheep who'll believe whatever sensationalist news story pops up on the feed.

[DrMikeNZ]

14 minutes ago, M.Yurizaki said:

This is not a temperature problem...

Not surprised. If it was solely a temperature problem, then more cards would have failed, and it would have been pretty clear in the early QC at EVGA.

The number of cards that have failed is small. I am still happy with my GTX 1080 FTW. I don't expect that it would fail, but like with all my computer components I have contingencies planned in case of failure.

[Arcmanov]

Unfortunately, despite installing the thermal mod pads, one of my 1070s bit the dust over the weekend.

EVGA was quick to approve the RMA though, so, there's that.

 

Switched to Gigabyte G1 1070s for the time being just in case though.

[Delicieuxz]

11 minutes ago, Arcmanov said:

Unfortunately, despite installing the thermal mod pads, one of my 1070s bit the dust over the weekend.

EVGA was quick to approve the RMA though, so, there's that.

 

Switched to Gigabyte G1 1070s for the time being just in case though.

What are you going to do with the extra GPUs when you get a replacement EVGA 1070?

[DrMikeNZ]

19 minutes ago, Arcmanov said:

Unfortunately, despite installing the thermal mod pads, one of my 1070s bit the dust over the weekend.

EVGA was quick to approve the RMA though, so, there's that.

Sorry to hear that. In what way did it bite the dust, flash of light and spectacular amount of smoke from the VRM, or something unrelated?

I never was convinced that the sub-optimal cooling was linked to the failures, at least the failure rate hasn't been extremely high.

11 minutes ago, Delicieuxz said:

What are you going to do with the extra GPUs when you get a replacement EVGA 1070?

You can never have too many GPUs.

[Arcmanov]

My system just shut off.  No warning.  Tried to reboot, system refused to power on.

I initially thought it was the PSU, but some troubleshooting found the culprit to be the primary 1070.

Complete and catastrophic power failure. Can't detect any burns or smoke though.

 

As for when I get the replacement...I'll keep it as a spare for now.  The other 1070 is in my HTPC. (If that one fails as well, I won't be surprised  )