RX 7900 XTX reference at possible vapor chamber design problem + very high hot spot (110c)

[AsRock]

It can't just be because the "vapor chamber" can't handle the heat load. When the 6900 XT uses the same type of heatsink & has about the same thermal heat load is watts when overclocked.

Maybe they over worked the machines that make these ( the molds or what ever ), there fore getting a bunch with weak seals or some thing. Then again you get AIO with contaminates so shit happens right.

 

[Vya Domus]

The AIB have pretty extensive knowledge of the chip, design, and schematics because be it AMD or Nvidia, they do share the documentation with their partners.

"Extensive"

No they don't, this is complete nonsense. They only tell them what they need to know to get the things working, no more no less. I don't see why the hell AMD or Nvidia would tell every single intricate detail about how their chips work.

They draw the schematics and the dies are produced by a third party, TSMC, Samsung, etc. So by extrapolating your little genius, unless TSMC, and Samsung tell them, they do not know either and everything is just a little secret!

You are aware everyone can technically look at the die of a chip and see everything for themselves, right ? Doesn't mean anything, reverse engineering chips according to just manufacturing schematics would be hell, so no, their secrets are still safe.

 

[ARF]

I also suspect that the partners don't get enough information because of the fear of leaks.
Of course, the result is what we see now - the partners don't know what to manufacture.

 

[SaLaDiN666]

Seems like you set a high bar with your first post.

And your still arguing about how cool your 3090 is while also clearly being aware that the number you see is decided by Nvidia not you, funny.

Not really on topic though, so you don't like AMD is my main takeaway of your POINT, and....

GPU's run hot, shocker.

A chip designed to run until it's limit works when it hits it's limit as expected, shocker.

A 3090 with a more expensive cooler fittied to it is cooled better than another card, well you did pay for it so again shocker.

Now please do get back to topic or spam some other thread with shite.

I'm actually interested in what's going on with the 7900XTX not your old expensive piece of shit

Yeah I went there troll.

yes, the numbers are clearly faked and Nvidia conspiracy, the card is secretly overheating and the internet is going after AMD only. It is also illegal but no class action suit like in the case of fake AMD 8 cores.

For example, the problem with the 6XXX series was so big that AMD had to publicly release a statement that 110C is ok, within specs and the cards thermal throttle by design despite it causing other issues..

Oh no, I invested my time and money into 6 different AMD GPUs in order to give them chance. I experienced exactly the same issue with all of them, this is the third generation exhibiting the same problem which is only getting progressively worse and worse but please, leave my AMD alone!

I remember the same treatment for Nvidia when their VRAM was overheating on their founder editions GPUs and what? It was justified. Deal it with.

 

[the54thvoid]

The OP is about the possible vapor chamber design problem. That is the topic of discussion.

Stay on topic, please - cooler design. Hell, how about someone talks about fluid mechanics and why it's possibly at fault?

 

[TheoneandonlyMrK]

yes, the numbers are clearly faked and Nvidia conspiracy, the card is secretly overheating and the internet is going after AMD only. It is also illegal but no class action suit like in the case of fake AMD 8 cores.

For example, the problem with the 6XXX series was so big that AMD had to publicly release a statement that 110C is ok, within specs and the cards thermal throttle by design despite it causing other issues..

Oh no, I invested my time and money into 6 different AMD GPUs in order to give them chance. I experienced exactly the same issue with all of them, this is the third generation exhibiting the same problem which is only getting progressively worse and worse but please, leave my AMD alone!

I remember the same treatment for Nvidia when their VRAM was overheating on their founder editions GPUs and what? It was justified. Deal it with.

Yes of course you did, just like I would buy 6 different Intel Dgpu to give them a chance, no I lied, I'd buy one, I'm normal.

@the54thvoid sorry I was typing when you were posting, but I did try minimum vinegar at least.

 

[delshay]

@ALL

If you have 110C hotspot you have to look at the "average delta difference" between the "core & hotspot temperature sensor". They are connected to the same coldplate so there should not be a big massive difference

For example am working on my Vega card & the hotspot is always 8C higher than the core, but at idle they are more or less the same temperature. Look at other users result to get an idea where your card should be.

 

[HD64G]

If the design was faulty, ALL of the GPUs with that design would have the same behaviour. So, it is a manufaturing process problem that allows a percentage to have faulty chambers.

 

[AusWolf]

yes, the numbers are clearly faked and Nvidia conspiracy, the card is secretly overheating and the internet is going after AMD only. It is also illegal but no class action suit like in the case of fake AMD 8 cores.

For example, the problem with the 6XXX series was so big that AMD had to publicly release a statement that 110C is ok, within specs and the cards thermal throttle by design despite it causing other issues..

Oh no, I invested my time and money into 6 different AMD GPUs in order to give them chance. I experienced exactly the same issue with all of them, this is the third generation exhibiting the same problem which is only getting progressively worse and worse but please, leave my AMD alone!

I remember the same treatment for Nvidia when their VRAM was overheating on their founder editions GPUs and what? It was justified. Deal it with.

Running a GPU up to its designed safe temperature limit is surely illegal, we should definitely sue AMD's ass over it! How dare they!

What about recent Intel CPUs? Or the single-fan EVGA 2060 that I bought for my brother that actually does run into a hard temperature limit under full load? Let's just make it simple and sue everyone!

In the meantime, it wouldn't hurt if you answered my questions regarding your issues above... or even better, return to the original topic, the 7900 XTX.

@ALL

If you have 110C hotspot you have to look at the "average delta difference" between the "core & hotspot temperature sensor". They are connected to the same coldplate so there should not be a big massive difference

For example am working on my Vega card & the hotspot is always 8C higher than the core, but at idle they are more or less the same temperature. Look at other users result to get an idea where your card should be.

As far as I know, Vega GPUs don't have hotspot temperature sensors. GPU-Z only reports a calculated difference, just like on Nvidia cards up to Ampere.

Edit: That's why it's always 8 °C - otherwise, the two values would be the same at idle, and there would only be a difference under load.

 

[Vya Domus]

@ALL

They are connected to the same coldplate so there should not be a big massive difference

Well that's the thing, there is a massive difference.

 

[Shrek]

One thing I thought was missing from the video was to let the card cool down and then show it recovers normal vertical operation.

 

[nguyen]

Probably when a particular part of the GCD put out too much heat, it disturb the water flow of the vapour chamber due to the uneven shape.

Der8eur is using a torture test that heat up that very specific part of the GCD, meanwhile different games can stress different area on the GCD

 

[delshay]

As far as I know, Vega GPUs don't have hotspot temperature sensors. GPU-Z only reports a calculated difference, just like on Nvidia cards up to Ampere.

Edit: That's why it's always 8 °C - otherwise, the two values would be the same at idle, and there would only be a difference under load.

Look at the delta difference between the core & the hotspot of my Vega card highlighted. Get the OP of the video in this thread to sacrifice warranty on one of his cards & add LM to the die. He probably won't do it as I don't think it's a vapor chamber problem.

Here's my Vega card. Look at the delta difference between core & hotspot. it goes up & down in tandem.

 

[Steevo]

Someone with a riser cable or with some grunt try flipping their card/whole case so the fans face up after it reaches the 110C and let us knownwhat happens. Reading other places too it seems that some vapor chambers were made with too little fluid, or a defect in wicking material causing this issue. If this is the case AMD will have to go after the heatsink vendor for faulty product.

 

[TheoneandonlyMrK]

Look at the delta difference between the core & the hotspot of my Vega card highlighted. Get the OP of the video in this thread to sacrifice warranty on one of his cards & add LM to the die. He probably won't do it as I don't think it's a vapor chamber problem.

Here's my Vega card. Look at the delta difference between core & hotspot. it goes up & down in tandem.


View attachment 277174

While I get what you are saying I think it worth mentioning that your and my Vega are not exactly the same as these newer card's, these have even more sensors and importantly considerably better power gating across the die, so the thermal deviation across the die can be higher.

It is beyond our knowledge where are the sensors and the details of the thermal control algorithm, that's a shame tbh.

 

[delshay]

Someone with a riser cable or with some grunt try flipping their card/whole case so the fans face up after it reaches the 110C and let us knownwhat happens. Reading other places too it seems that some vapor chambers were made with too little fluid, or a defect in wicking material causing this issue. If this is the case AMD will have to go after the heatsink vendor for faulty product.

Let me say this again. If it's a vapor chamber problem then the GPU temperature should also rise. We need to find out what is the minimum delta difference between core & hotspot. The only way I know is to add LM on the die, unless someone knows another method.

 

[eidairaman1]

Hey there's some Sense in this link.

AMD Radeon RX 7900 XTX May Feature Faulty Coolers, Causing Overheating

Is the 7900XT affected at all? Mine seems fine but I’m curious as they feature a similar but not identical cooler. I mean Derbauer cant be 100% sure either, he has a handful of cards of which some were faulty, some where not, we dont really know on what scale this really is a problem and we...

www.techpowerup.com