• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Nvidia RTX 4090 power connectors melting?

ZeroBarrier said:
Here, let me tl;dr as easily as I can.

View attachment 684244

Still a nothing burger. The same as it was when it was finally discovered that the one thing in common was user error.
Click to expand...
That's an interesting way to interpret this. 12VHPWR has been around for 1 gen. So in 2 years you have 4% failure. 6-pin and later, 8-pin started what....20 years ago? Not a great rate in 2 years of usage for 12VHPWR. Hardly the same thing and not great looking for 2 years of 12VHPWR vs 20 years of 6-pin and 8-pin PCI-E.

Now I will agree with others that this is simply a self-reporting poll, so hardly a real firm scientific statistic. But still.

Also at a certain point a design that makes it too easy to have "user error" is the fault of the design if enough users have issues. As also reported in the video, plenty of the card designs made if incredibly difficult to fully seat the cable within say a recessed plug in the cooler, or other things where verifying it was in all the way was not easy to do. And personally I'd say if you are plugging a cable in and still have to question if its in all the way even though it feels like it is and is even "clicked in" then that's a fault of the design. Now are there genuine user errors of people not using the cable correctly? Absolutely, I won't claim there isn't. Just saying I don't understand the argument it can only be one or the other. It's probably both. A better retention design would eliminate a lot of this "user error".
 
Last edited:
Sir Beregond said:
That's an interesting way to interpret this. 12VHPWR has been around for 1 gen. So in 2 years you have 4% failure. 6-pin and later, 8-pin started what....20 years ago? Not a great rate in 2 years of usage for 12VHPWR. Hardly the same thing and not great looking for 2 years of 12VHPWR vs 20 years of 6-pin and 8-pin PCI-E.

Now I will agree with others that this is simply a self-reporting poll, so hardly a real firm scientific statistic. But still.

Also at a certain point a design that makes it too easy to have "user error" is the fault of the design if enough users have issues. As also reported in the video, plenty of the card designs made if incredibly difficult to fully seat the cable within say a recessed plug in the cooler, or other things where verifying it was in all the way was not easy to do. And personally I'd say if you are plugging a cable in and still have to question if its in all the way even though it feels like it is and is even "clicked in" then that's a fault of the design. Now are there genuine user errors of people not using the cable correctly? Absolutely, I won't claim there isn't. Just saying I don't understand the argument it can only be one or the other. It's probably both. A better retention design would eliminate a lot of this "user error".
Click to expand...
There is no interpretation here. Facts are facts, and the facts are clear that while the connector could be better, and no one disagrees with this assertion, the connector isn't the reason for all the drama. By and large (with the exception of the shoddy third-party connectors), the melting connectors were caused by user error.

You know what else we needed to change because the design allowed for user error? Disposable coffee cups. Literally have to remind people that have less than 2 brains cells to rub together to form and intelligent thought that the hot contents that they ordered are HOT. Maybe that's what Nvidia should do, put a warning on their GPU to remind those that can't critically think to seat the connector all the way in and make sure it's seated properly before proceeding.

It can absolutely be more than just one way or the other, but the facts remain that the only way anyone has gotten the connector to fail is by intentionally not seating it correctly. It's the sole reason it took as long as it did for anyone to figure out what the issue could be. Now I know not everyone deals with Joe public en mass, but let me tell you as someone who deals with hundreds of people daily, the vast majority of people are so dumb that I wouldn't trust a single one of them with a potato gun.
 
ZeroBarrier said:
There is no interpretation here. Facts are facts, and the facts are clear that while the connector could be better, and no one disagrees with this assertion, the connector isn't the reason for all the drama. By and large (with the exception of the shoddy third-party connectors), the melting connectors were caused by user error.

You know what else we needed to change because the design allowed for user error? Disposable coffee cups. Literally have to remind people that have less than 2 brains cells to rub together to form and intelligent thought that the hot contents that they ordered are HOT. Maybe that's what Nvidia should do, put a warning on their GPU to remind those that can't critically think to seat the connector all the way in and make sure it's seated properly before proceeding.

It can absolutely be more than just one way or the other, but the facts remain that the only way anyone has gotten the connector to fail is by intentionally not seating it correctly. It's the sole reason it took as long as it did for anyone to figure out what the issue could be. Now I know not everyone deals with Joe public en mass, but let me tell you as someone who deals with hundreds of people daily, the vast majority of people are so dumb that I wouldn't trust a single one of them with a potato gun.
Click to expand...
Yes interpretation because time a product exists definitely matters in terms of the rate of reported errors. 2 years vs 20 years says something. But please go ahead and continue to pretend 4% in 2 years is fine and "the same thing".

Again you make it an either or while in passing saying the design could be better. At the end of the day if the same person who has been using 6-pin and 8-pin forever without issue and definitely knows how to plug a cable in suddenly runs into a melting connector on a 12VHPWR because it still melted after technically being "clicked in", I still go to, well the root cause then is the design. If its not seated all the way, why is it "clicked in".

Now that's different than the users who genuinely don't pay attention and yes there is certainly a "user error" factor here.

Maybe I am alone in this, but instead of "picking sides" like it's some weird hill to die on, I can see where both possibilities exist. Yes...user error, yes there are idiots, but also this design needs work and the ease at which a design can allow user error is a factor in some of these cases. Nuance is a thing. You just don't see the same thing with people posting about their melted 6-pin and 8-pin connectors like you did with 12VHPWR. So either all the idiots only showed up in the past 2 years, or maybe the design needs some work.
 
Last edited:

Hero cat saves owner as PC starts burning from NVIDIA RTX 4090 meltdown​

byMichael Gwilliam
Published: Apr 07, 2026, 23:23

A PC owner has credited their cat with saving them from a dangerous situation after their NVIDIA RTX 4090 began to melt and smoke while unattended.

According to reports, the incident started when the user stepped away from their computer, only to hear their cat meowing unusually loudly from another room.

When they returned, they discovered smoke coming from their PC along with a strong burning smell, caused by a melted GPU power connector.

The owner quickly unplugged the system to prevent further damage after noticing the PC wouldn’t shut down normally, potentially avoiding a more serious fire.

Cat saves NVIDIA RTX 4090 owner’s life​

The story originally surfaced on a Taiwanese forum similar to Reddit, before being picked up by tech outlets.

https://www.dexerto.com/entertainme...s-fire-from-nvidia-rtx-4090-meltdown-3348004/
 
Marees said:

Hero cat saves owner as PC starts burning from NVIDIA RTX 4090 meltdown​

byMichael Gwilliam
Published: Apr 07, 2026, 23:23

A PC owner has credited their cat with saving them from a dangerous situation after their NVIDIA RTX 4090 began to melt and smoke while unattended.

According to reports, the incident started when the user stepped away from their computer, only to hear their cat meowing unusually loudly from another room.

When they returned, they discovered smoke coming from their PC along with a strong burning smell, caused by a melted GPU power connector.

The owner quickly unplugged the system to prevent further damage after noticing the PC wouldn’t shut down normally, potentially avoiding a more serious fire.

Cat saves NVIDIA RTX 4090 owner’s life​

The story originally surfaced on a Taiwanese forum similar to Reddit, before being picked up by tech outlets.

https://www.dexerto.com/entertainme...s-fire-from-nvidia-rtx-4090-meltdown-3348004/
Click to expand...
I just replaced two power cords that I bought at Walmart the cords were maybe 50+ years old. I have Power Strips for every outlet to prevent cording hanging out of the Wall.
I would like the get a UPS Battery Power thingy but alot of them don't have enough outlets and are kinda bulky I haven't really needed once since 1996 so I figure it would be a waste.
 
ASUS ROG Equalizer Targets Safer 12V-2x6 GPU Power Delivery

IMG_4587.jpeg
 
Anything but going back to PCIe power connector and actually making it 4 voltage + 4 ground pins and using 3 of them.
 
I've taken several thermal images of mine and never had a heat issue with the connector.

Use common sense, don't put pulling stress on the connector, you will be fine.
 
GoodBoy said:
I've taken several thermal images of mine and never had a heat issue with the connector.

Use common sense, don't put pulling stress on the connector, you will be fine.
Click to expand...

There are engineers on Youtube putting ammeters on the wires showing a huge disparity of current on various wires during operation. This was never just about connector stress. I'm sure there are various examples where there are no issues, but the inconsistency necessitating this product shows manufacturers are recognizing this problem.

And it's not limited to Nvidia. Sapphire's now had several failures on their 9070XT model using the connector as well. AMD fans are just not as interested in defending it because AMD board partners are allowed to make 8-pin connector options.
 
GoodBoy said:
I've taken several thermal images of mine and never had a heat issue with the connector.

Use common sense, don't put pulling stress on the connector, you will be fine.
Click to expand...
That and don’t use cheap suspect aftermarket cables just to achieve a “good look”.
 
HockeyJon said:
There are engineers on Youtube putting ammeters on the wires showing a huge disparity of current on various wires during operation. This was never just about connector stress. I'm sure there are various examples where there are no issues, but the inconsistency necessitating this product shows manufacturers are recognizing this problem.

And it's not limited to Nvidia. Sapphire's now had several failures on their 9070XT model using the connector as well. AMD fans are just not as interested in defending it because AMD board partners are allowed to make 8-pin connector options.
Click to expand...
Oh the connector could be better no doubt. More idiot proof, since it is so high current.
There are examples of the 8-pin connector's melting as well.
So, again, be careful, it will be fine.
I like German Muscles' suggested connectors, those look stout. Wires would be stiffer... but capable.
 
GoodBoy said:
Oh the connector could be better no doubt. More idiot proof, since it is so high current.
There are examples of the 8-pin connector's melting as well.
So, again, be careful, it will be fine.
I like German Muscles' suggested connectors, those look stout. Wires would be stiffer... but capable.
Click to expand...

The point is that the connector is not the only problem here. They’re redlining its capacity and the current distribution often leaves a lot to be desired. Even if you’re super careful connecting it, it’s still a bad solution until they fix that problem.
 
You must log in or register to reply here.
Back
Top