News Defective Vapor Chamber May Be Causing RX 7900 XTX Overheating Issue

[Eximo]

If the sintering only exist around the hot plate then that by definition is not a vapor chamber. The cold plate side needs sintering back to the hot plate to allow capillary action to work.

Sintered copper around the 9 posts above the GPU. Copper mesh everywhere else top and bottom of the vapor chamber.

Not sure I understand the design myself. Perhaps the mesh is preferred over such a large area for ease of construction. As pointed out in the video, they just squished the mesh under the structural supports.

 

[zx128k]

These designs should always fail safe. A fuse should always blow before a connector or wire melts. Preferable (but much harder to design) would be that no current flows until the connector is fully seated.
All bets would be off if there were major accidental damage, like somebody spilling a soda on the gpu... But there's no excuse for unsafe design when we're talking about such a minor and easily missed user error.

Thats not how it really works. The user is ment to install the power cable correctly. There are lots of electronics that will break or go on fire if the power is not installed correctly. This issue happens with micro USB cables. There have been recalls if it was the cable that was the problem. Extension cords can overheat and cause fires when used improperly.

If you connect the circuit wires to the wrong terminals on an outlet, the outlet will still work, but the polarity will be backward. When this happens, a lamp, for example, will have its bulb socket sleeve energized rather than the little tab inside the socket. You can be killed by this issue.

Most if not all electricity wiring issues are dangerous. The end user has the responsibility to do it correctly and safely.

With PC's drawing to much power from the motherboard can melt the 12 volt pins on the 24 pin atx connector. GPU's didn't limit the power drawn in the past from the PCIe socket.
Melted 8 pin cpu connector (multiple pins) and 24 pin connector (just single pin)

Melted 8 Pin picture.
Melted 24-pin picture.

A faulty PSU can also cause this issue.

Burnt 24-Pin connector and random shutdowns Reason here, It's a bad wire to socket connection I suppose and it's a good idea to get a new PSU before you break something... I was told it was either a bad connection or my power supply just overheated and over-volted on the pins causing the first bsod. People can do this easly by overclocking a CPU or with sli setups. This is why getting enough power is more than a good idea. Why overclocking old systems with old PSUs can end in a burning smell.

The whole core attack on nVidia is that this melting happens only with the new 12-pin gpu power connector and not the old 8-pin.
Here's an 8-pin GPU connector burned.

The RX 580 would burn up 8-pin power cables as well.

Here is another 8-pin user error. learned this in the hard way. 3090.. always use 2 seprate pcie power for pin1 and 2.

Yet its just the 12-pin connect that has design issues or maybe its a PR attack on nvidia? As this issue has been happening on other connectors whenever a user does something dumb. Even 8-pin connector burn when used wrong.

 

[zx128k]

Turing gpus and above have limits to power draw on each 8-pin and fuses. Still people use spliters and use one 8-pin for two sockets on the gpu. Burning the 8-pin connector. Couldn't add this to last post. Server error.

 

[tamalero]

LOL like Jobs holding the phone the wrong way IPhone 4 comment?

Making sure the cable is fully plugged in solved the Nvidia issue.

Buying AIBS is not "holding the iphone the wrong way". They are unaffected, only AMD models with ref designs are the issue. Where is the lie?

As per Nvidia, Yes obviously, still affects all cards.

I still laughed how the whole Nvidia issue was self made.
Because the burning thing made everyone paranoic, they were unplugging and plugging their connectors again and again to check.
And they either slowly damaged the plug.. or not connected it fully, causing a self fulfilled prophecy of failure.


Still, the Nvidia plug has a very bad design.
It does not click when you expect it would. My Gigabyte 4090 feels extremely tight and even if its not plugged all the way in. I feared I could bend the connector.
Thankfully I managed to plug it all in.

What they need to do? just modify the receiving plug (the female portion) to make sure it clicks when all plug.
So yes, its also a design issue.

 

[tamalero]

No kidding. A defective vapor chamber is a bit more severe than not connecting a cable securely. Jeez.

How so?
A cable not connecting securely can lead to a fire.

A defective vapor chamber can lead to just the card thermally throttling. 😵

Why didn't anyone at AMD even think to test the Card with fans down? That is the most common position for the card to be in. We don't all have test benches you know...

Buying an AIB non reference XT at least let me avoid this issue in 2 ways. Its not XT (yet) and its only reference cards

The reference tested designs (as mentioned by DerBauer), Might have been from a different manufacturer. Aka the failure was not on the first boards, but only on the retail versions.

So imho, perhaps its not a design issue but a quality issue.

 

[zx128k]

Rumor mill, upto 11% (? watch the video) of cards could be affect and AMD rumored to have not enough cards to replace the faulty ones.