navi 31 AMD Declines Radeon RX 7900 XTX RMA For Hitting 110C Junction Temps, Says “Temperatures Are Normal” (Updated #2)

[NumLock21]

TJunction of 110C has been within acceptable specs for a few years, back then AMD reference cards were blower style one, but these new ones have triple fans just like AIB cards, so bad or faulty cooler design?

[BiG StroOnZ]

Okay, so it seems that according to the updated Reddit post, after making a stink, that the Redditor in question was finally accepted for a return (along with a few others):

 

 

It also appears that the support line for the storefront is not ran by AMD, but by a company known as Digital River. Will add to OP. 

[Jada88]

On 12/28/2022 at 12:53 PM, BiG StroOnZ said:

Okay, so it seems that according to the updated Reddit post, after making a stink, that the Redditor in question was finally accepted for a return (along with a few others):

 

 

It also appears that the support line for the storefront is not ran by AMD, but by a company known as Digital River. Will add to OP. 

AMD lied. AMD is the one who handles the RMA process and there is no way they would have not known about this issue that has been blowing up for 2 weeks, nor could a 3rd party simply decline RMAs off no criteria as that criteria had to come from somewhere. This is especially true for a 110+c issue which is not a reasonable decline of RMA.

 

AMD just tried to throw Digital River under the bus like they've done in the past to defer blame for the situation. Proof here:

In response to the bug, Digital River told PCMag it actually doesn't host AMD's online store. “AMD’s site is utilizing our global seller services for managing payments, taxes, fraud and compliance. We are the seller of record, which is why Digital River’s name appears on the transaction but we do not host their store.”

https://www.pcmag.com/news/bug-in-amds-online-store-allowed-people-to-easily-buy-graphics-cards

There is no way AMD did not catch this massive hardware flaw that impacts 100% of all reference models, due to being a hardware level design and manufacturing flaw, or missed the highly fluctuating clock rates during testing before selling these GPUs. There is also no reasonable way AMD is blind to such vocal complaints only a week or two after their major product released.

 

AMD is just playing its usual games.

[starsmine]

Der8auer appears to have done enough testing to be pretty confident in his conclusion of MBA cooler cards having a flaw in the vapor chamber QC and or design. If validated by other testing, I think he is correct in his call to action of these need a recall, because unlike the 12VHPWR connector issue which was completely user error. This issue is not. Majority of cases are not test benches and are going to be mounted horizontally. 



So buy non reference coolers, or water-cool the reference. 

 

AMD is correct in terms that the chips are designed for a TJMAX of 110C. a card getting there is not the issue, and is a misleading headline imo. That part is still inside spec. But that the vapor chamber running dry causing the card to throttle rather then hold a steady boost is the actual issue. 

[WereCat]

6 hours ago, starsmine said:

Der8auer appears to have done enough testing to be pretty confident in his conclusion of MBA cooler cards having a flaw in the vapor chamber QC and or design. If validated by other testing, I think he is correct in his call to action of these need a recall, because unlike the 12VHPWR connector issue which was completely user error. This issue is not. Majority of cases are not test benches and are going to be mounted horizontally. 



So buy non reference coolers, or water-cool the reference. 

 

AMD is correct in terms that the chips are designed for a TJMAX of 110C. a card getting there is not the issue, and is a misleading headline imo. That part is still inside spec. But that the vapor chamber running dry causing the card to throttle rather then hold a steady boost is the actual issue. 

Saying that TJmax at 110C is fine completely changes based on context.

Technically AMD is right by saying it's fine but if the card reaches 110C and is thermal throttling and you get bad performance then it's not fine.

 

Obviously the protections on the card don't allow it to go past 110C and the card would probably shut down rather than cross this target if it got even more out of hand so it's not like you can RMA it because it got even hotter as the card will rather downclock a lot to no exceed 110C... so nice excuse from AMD side to just say that 110C is fine.

 

Although they at least started to accept RMA but still charge for shipping, lol. I wonder if there will be a big pushback once this blows up even more after this video.

 

EDIT:
I find it interesting how the NVIDIA cable "issue" blowed up so fast and seemingly nobody even cares about this one, lol.

[HenrySalayne]

7 hours ago, WereCat said:

EDIT:
I find it interesting how the NVIDIA cable "issue" blowed up so fast and seemingly nobody even cares about this one, lol.

Because your card throtteling will not burn your house down?

 

9 hours ago, starsmine said:

Der8auer appears to have done enough testing to be pretty confident in his conclusion of MBA cooler cards having a flaw in the vapor chamber QC and or design. If validated by other testing, I think he is correct in his call to action of these need a recall, because unlike the 12VHPWR connector issue which was completely user error. This issue is not. Majority of cases are not test benches and are going to be mounted horizontally. 

It's definitely an issue and I agree with Der8auer that AMD has to address this problem ASAP. I don't think a total recall of cards is necessary. The cooler seems to work fine for the majority of buyers and why should they put these costumeres through the trouble of replacing their GPU?

AMD should

- halt sales.

- make RMAs a straightforward process for costumers with affected GPUs.

- make tools and guides for costumers to check if they are affected.

- allow straightforward refunds for unsatisfied customers.

- not take month to admit this and start the above process. 

[porina]

8 hours ago, WereCat said:

I find it interesting how the NVIDIA cable "issue" blowed up so fast and seemingly nobody even cares about this one, lol.

Too new? Or many of those making the most noise are AMD fanboys.

 

35 minutes ago, HenrySalayne said:

Because your card throtteling will not burn your house down?

How many houses have burnt down from power cable related problems?

 

35 minutes ago, HenrySalayne said:

I don't think a total recall of cards is necessary. The cooler seems to work fine for the majority of buyers and why should they put these costumeres through the trouble of replacing their GPU?

I'd agree there doesn't need to be a forced recall. If the current theory about the vapour chamber is proven, which it has yet to be, then certainly anyone with an affected device should be up for a no fuss replacement. The concern I'd have is for those that might have a "bad" cooler but don't know it yet since they use it in such a way it isn't obvious.

 

I also have to wonder if examining the vapour chamber is within Der8auer's capability. Anyone can open it up to have a look inside, but to check it has the right amount and type of coolant for the job is another level. Could it be "refilled"? But this is then getting more into design like considerations.

[HenrySalayne]

50 minutes ago, porina said:

How many houses have burnt down from power cable related problems?

There have been no fires reported from NZXT H1s lighting up, but it's a risk. Fire needs three things: flammable material, oxygen and a ignition source. You will find the first two in many PC cases and the surrounding area; adding an ignition is a valid concern.

 

56 minutes ago, porina said:

I also have to wonder if examining the vapour chamber is within Der8auer's capability. Anyone can open it up to have a look inside, but to check it has the right amount and type of coolant for the job is another level. Could it be "refilled"? But this is then getting more into design like considerations.

These thing belong in the hand of engineers who know what they are doing and who have access to a lab equipped for this kind of testing. There is not much Roman or other knowledgeable people with good skills and a workshop could inspect. It`s not only about the (amount of) coolant / refrigerant (because the medium undergoes a phase change which would put it in the latter category), but also the (vapour) pressure, the structure of the porous material deposited on the inside and the thermal flux in the case.

Vapour chambers as well as heat pipes can be overloaded and stop working after reaching a critical temperature. And this is probably what Der8auer is showing at the end of his video with tilting the cards. They start to throttle and tilting them back doesn't resolve the issue.

[WereCat]

1 hour ago, HenrySalayne said:

Because your card throtteling will not burn your house down?

While fire is a huge risk there is still an issue with a product that costs $1k+ and everyone and their grandma was an expert on PCIe power cables when that happened. I just find that amusing that AMD keeps getting a pass way more often for some reason. Maybe I'm just saying this too soon though. 

[vertigo220]

3 minutes ago, WereCat said:

While fire is a huge risk there is still an issue with a product that costs $1k+ and everyone and their grandma was an expert on PCIe power cables when that happened. I just find that amusing that AMD keeps getting a pass way more often for some reason. Maybe I'm just saying this too soon though. 

I don't see how they're getting a pass. When the Nvidia issue was discovered, I read about it on tech sites and there were a bunch of people commenting. Since this issue was discovered, I've read about it on tech sites and there have been a bunch of people commenting. Both instances had their share of people calling out Nvidia/AMD as well as people saying it's not a big deal and Nvidia/AMD aren't to blame. I don't really see much difference personally, but maybe that's just me.

[leadeater]

9 hours ago, WereCat said:

I find it interesting how the NVIDIA cable "issue" blowed up so fast and seemingly nobody even cares about this one, lol.

Apparently 86% of the market is Nvidia so that's a rather good reason why. X product nobody really has has an issue, nobody continued to care lol.

 

But I guess a large part was one issue had melting connectors and potential fire and the other does not, with a dose of "New annoying connector I don't think we needed" with a sprinkle of "I already dislike Nvidia so lets have at".

[leadeater]

2 hours ago, HenrySalayne said:

It's definitely an issue and I agree with Der8auer that AMD has to address this problem ASAP. I don't think a total recall of cards is necessary. The cooler seems to work fine for the majority of buyers and why should they put these costumeres through the trouble of replacing their GPU?

A company can issue a recall and you don't actually have to send your card back, and they still have to honor the warranty. So if it's working fine then you are actually within your right to not send the card back. Not all recalls are mandatory, Gov ones issued for safety generally are. 

 

2 hours ago, HenrySalayne said:

AMD should

- halt sales.

- make RMAs a straightforward process for costumers with affected GPUs.

- make tools and guides for costumers to check if they are affected.

- allow straightforward refunds for unsatisfied customers.

- not take month to admit this and start the above process. 

That's essentially what a manufacturer recall notice is.

[porina]

26 minutes ago, HenrySalayne said:

There is not much Roman or other knowledgeable people with good skills and a workshop could inspect.

At the end of the vid he said words to effect it could be a follow up, so let's see where it goes from there.

 

Thinking more about the recall potential, I feel like we're still missing a part. Do we have demonstrably "good" samples of this cooler that remain working as expected after going through the test process? I wonder how many are actually affected? The appropriate path will vary with that.

 

In light of that video, I'm now wondering if there are owners of these GPUs turning their cases over to see if they're affected.  

[leadeater]

3 minutes ago, porina said:

At the end of the vid he said words to effect it could be a follow up, so let's see where it goes from there.

 

Thinking more about the recall potential, I feel like we're still missing a part. Do we have demonstrably "good" samples of this cooler that remain working as expected after going through the test process? I wonder how many are actually affected? The appropriate path will vary with that.

 

In light of that video, I'm now wondering if there are owners of these GPUs turning their cases over to see if they're affected.  

I think it's likely at a point where AMD need to do their own investigation to find out if it's a particular batch of GPUs. May or may not be a design issue but if it's just a manufacturing issue then it could well be isolated to a batch of GPUs, although that number is likely quite significantly high as they would be getting made in large bulk for launch.

[Forbidden Wafer]

12 hours ago, starsmine said:

Der8auer appears to have done enough testing to be pretty confident in his conclusion of MBA cooler cards having a flaw in the vapor chamber QC and or design. If validated by other testing, I think he is correct in his call to action of these need a recall, because unlike the 12VHPWR connector issue which was completely user error. This issue is not. Majority of cases are not test benches and are going to be mounted horizontally. 



So buy non reference coolers, or water-cool the reference. 

 

AMD is correct in terms that the chips are designed for a TJMAX of 110C. a card getting there is not the issue, and is a misleading headline imo. That part is still inside spec. But that the vapor chamber running dry causing the card to throttle rather then hold a steady boost is the actual issue. 

[dogwitch]

we know who ever manf said bad batch of cooler... said company just lost a client.

[CommanderAlex]

1 hour ago, WereCat said:

While fire is a huge risk there is still an issue with a product that costs $1k+ and everyone and their grandma was an expert on PCIe power cables when that happened. I just find that amusing that AMD keeps getting a pass way more often for some reason. Maybe I'm just saying this too soon though. 

Probably cause everyone cheers on the underdog. 

[AlwaysFSX]

We'd start up the AMD meme presses but they've been on fire since the HD7970.

 

Whoever designed this vapor chamber fuuuuuuuuuuucked it. How unfortunate.

[leadeater]

2 hours ago, AlwaysFSX said:

We'd start up the AMD meme presses but they've been on fire since the HD7970.

HD 4000 series even. My 4850 ran hot AF and so did my 6970.

[Mark Kaine]

19 hours ago, WereCat said:

I find it interesting how the NVIDIA cable "issue" blowed up so fast and seemingly nobody even cares about this one, lol.

Amd gpu department is a lot smaller than nvidia afaik. sure amd does a lot of gpus but numbers pale in comparison there are just more nvidia users and nvidia also has this "can do no wrong" aura around them so of course its kinda a bigger deal, even though this flaw is actually worse than the cable issue. 

 

personal opinion: i find it remarkable whenever amd has the chance to beat nvidia they have a lot of sudden issues,  happened also with the hbm vega cards for example,  they had a really good chance at the crown and then somehow decided to not make many of them.

imagine if amd would flood the market with high quality, reasonable priced cards right now.  But nope? issues after issues and "weird" pricing/ value...

 

btw im not implying anything specific, just saying that its remarkable and noticeable. 

[AlwaysFSX]

15 hours ago, leadeater said:

HD 4000 series even. My 4850 ran hot AF and so did my 6970.

My brain starts deleting stuff after about a decade lol

[Quackers101]

4 hours ago, Schnoz said:

AMD right now:

When team red asked for a tube of water

Spoiler

dont go here

Spoiler

 

[Senzelian]

That's an unfortunate start for a otherwise good GPU generation.

[porina]

He opened it up. Don't think it progresses this issue but I found it interesting regardless.

 

[Quackers101]

5 hours ago, porina said:

He opened it up. Don't think it progresses this issue but I found it interesting regardless.

dam, hope it was tested alone in how it handles x amount of wattage of heat. and compared to another.