News RTX 4090's 16-Pin Connector Melted After One Year of Usage

[mac_angel]

In wonder if the guy frequently/infrequently removed and inspected the cable thus increasing chance of failure. Paying for one of these would make anyone nervous, poor guy. I just hope people with OG 4090s don't suddenly start checking, just leave it.

I check mine often because I get nervous about it, too. But I don't unplug it and plug it back in to check, I just look at it and maybe give it a little push to make sure it's all the way in and not slipping out for some reason. I'd rather be diligent than have to deal with a warranty argument and be without a GPU for several weeks.

 

[B0omSLanG]

I wonder if Corsair's 12VHPWR GPU Power Bridge will help with this. I just got a 4090 FE and should have the bridge today. At the very least, I'd hope it wouldn't burn the card itself. This is not a promo. It's just an alternative solution I found and bought along with a large Shift PSU.

 

[pf100]

I have a 4090 I bought used, so most likely no warranty, and I bought a new pcie5 power supply so it's a new connector from the psu plugged into it. I haven't decided yet exactly what I'm going to do to protect myself. I'll either buy a redesigned cablemod 180 degree extension adapter or if I just leave it as-is and the plug melts, I know how to solder and can remove the connector from the 4090, then use the 'four 8-pin' to 12vhpwr adapter that came with the card, cut off the 12vhpwr end, and solder the wires directly to the 4090's board. Either way, I'm not going to let a melted connector stop me.

 

[Avro Arrow]

Well, what do they expect, quality? They only cost $1600USD!

At times like this, I'm so glad that Radeons just use the tried-and-true standard PCIe Supplemental Power connectors. It makes me immune to Jensen's BS.

They really ought to be forced to do a recall on the 4090. The ongoing issues and knowing that it's 100% the fault of the connector, it's not fair to the consumers to have to foot the enormous bill for a replacement in a year because of bad engineering.

I wouldn't hold my breath...

When electrical terminals run hot due to being undersized or having poor physical contact they oxidize which increases the electrical resistance which creates more heat until the connector eventually melts.

Sure... but somehow this doesn't happen with any other connector in the entire PC. I have never in my life had a connector of any kind melt in my PC.

Nah. We all know nVidia and all of its most fervent followers said it's not an issue. People are just connecting the thing wrong. Fake news all the way. Yes. Definitely.

...

Bad sarcasm aside and on a more serious note, I hope nVidia foots the bill for everyone affected in either a direct or indirect way (via AiBs). That's one really good way to use a VERY small portion of their humongous margins of the AI-shekles they've been getting as of late. It would also gain good faith from an already disappointed enthusiast crowd feeling the milking to the bones.

Regards.

I must disagree with you. I thought that it was very GOOD sarcasm!

I'm just now in the market to complete my year-long build this Christmas and have enough saved for a 4090.... what is needed is a way to tell the older cards equipped with the 12VHPWR connect from the newer ones with 12V-2x6.... especially before I plunk my money down! 🤣

I think that what is needed is for people to not pay that much for a video card because that's the whole reason that they're at their current prices to begin with. I don't believe that there's any way to know which is the older connector. If you saved up for a year to get this, do yourself a favour and don't get this because you're the kind of person that it would hurt most if it were to happen. Also note that nVidia has been trying to blame "user error" even though said "user error" managed to not affect any other power connector in those users' PCs.

Honestly, after all that's happened and how nVidia has handled it, I'm kinda shocked that anyone would still be scrounging money together for that long to get a video card that's as overpriced (and seemingly unreliable) as the RTX 4090.

Do we know if any other RTX 40 series GPUs are affected? I purchased an ASUS TUF 4070 Ti back in June and a Corsair Premium 600W PCIe 5.0 / Gen 5 12VHPWR PSU Cable because it just looks better than that stock cable that comes with these GPUs. I don't recall seeing any 4070 Tis being affected, but that's not to say there haven't been any.

Yeah, but let's be honest, if it was happening to the RTX 4070 Ti and RTX 4080, we'd have heard about it and it wouldn't be an RTX 4090 issue, it would be an nVidia issue. I think that it has only been happening to the RTX 4090.

Having had many thousand system-integrator-assembled systems pass through my hands, and found issues far more egregious than "a cable was plugged most of the way in but not all the way in" (e.g. HSF loose and bouncing around inside chassis, because holddown screws were not present), user error is still far and away the most probable cause.
Particularly for 'experienced' system builders/integrators, who are clearly double-plus-good at their jobs and have never had any problems plugging a cable in before, so any post-inspection is clearly a total waste of time and unnecessary because of course it must be some other problem.

Maybe, but if that were the case, we'd be seeing electrical connectors of all kinds melting. We'd be seeing motherboard mains melting, CPU connectors melting and standard PCIe Supplementary Power connectors melting, but we're not. Either you know how to put two electrical connectors together or you don't, it's just that simple.

If only this connector is melting, then the problem has to be the connector because user error would be a common factor across all electrical connections, not just this one.

 

[Viking2121]

My buddies Asus strix 4090 burnt up a cable mod 180 degree adapter recently.

The GPU was vertically mounted no tension on the wires, and everything was fully seated and clipped in, after all the issues, it was thing he and I made sure was securely plugged in with no excess wight on the wire or weird pressure, card still cooked it.

It started playing Diablo 4, his PC would just black screen, but not crash, we could still communicate, and he could even still hear the game, it did a few times a month, He was playing another game and it was doing back to back, tired bios updates on the board and card, reinstalled the drivers with DDU, Nothing helped.

Finally we decided to remove the riser cable and just install the card the normal way, couldn't remove the plug from the card, took some force and got the plug out, sure enough had a few pins a bit chard and swollen and one even broke the plastic when it was forcibly removed. The plug on the card surprising didn't take much damage and was able to use it again, though cable mod did take care of him so no big issue there.

But when he bought it, over 2 grand, that just shouldn't happen, Plus I've been watching Northwestrepair on youtube, has been getting a lot of 4090s that's cracked from the weight of the cooler and im sure some shobby quality PCB's, Amazing how much of a mess this card is.

 

[edzieba]

Maybe, but if that were the case, we'd be seeing electrical connectors of all kinds melting. We'd be seeing motherboard mains melting, CPU connectors melting and standard PCIe Supplementary Power connectors melting, but we're not.

Well, we do. You can do a quick google for "PCIe 8 pin melted" or "Molex melted" or "EPS12V melted" and so on and find plenty of photos of other connectors that have melted.

 

[Avro Arrow]

Well, we do. You can do a quick google for "PCIe 8 pin melted" or "Molex melted" or "EPS12V melted" and so on and find plenty of photos of other connectors that have melted.

False equivalency. If power connectors of all kinds melted nearly as commonly as the 12VHPWR, the reaction to it happening wouldn't have been what it was. People would've said "Yeah, that kinda stuff happens all the time." but that's not what happened. Show me an article that details an issue surrounding standard PCI-Express Supplementary Power Connectors that warns about them melting if you don't do <insert instruction here> and then you might have a leg to stand on. Looking at pictures of melted connectors or reading forums means nothing because if it were a widespread issue like the 12VHPWR situation, it would have more than just pics and forum posts, it would have articles.

Something like this only becomes an issue if it happens frequently enough. There were never any warnings on standard PCI-Express Supplementary Power Connectors about them being prone to melting on any card. It sure wasn't because there weren't cards that didn't draw what the RTX 4090 does back in the day. The RTX 4090's TDP of 450W seems high to noobies but those of us who have been around for awhile know that the R9 295x2 had a TDP of 500W and the PowerColor Devil 13 Dual Core R9 290X had a TDP of 580W. If cards that came out ten years ago, cards that used more power than the RTX 4090, didn't have these problems, then there's no excuse for these problems with the RTX 4090.

There were still no issues with their PCI-Express Supplementary Power Connectors that ever saw the light of day in the tech press so I'm afraid that your argument falls flat.

 

[SyCoREAPER]

Personally I'm not concerned. As mentioned this is statistically low, nobody else has reporter melting since the whole "you're doing it wrong statements" about plugging it in all the way.

It's a butthole design that we are stuck with. I out of fear of obsolescence bought a 600w 12VHPWR to 4x8 cable. It's inevitable that 12VHPWR connector will die or be replaced. When that happened finding power supplies with native 12VHPWR will be slim pickings, slimmer than now.

I think this whole situation is worth being aware of but not fearmongering over.

 

[LabRat 891]

Just wait until next year when we see "How to spot original 12VHP RTX 4090 cards on eBay before buying" articles.

I sure hope so.
It'd be nice to get a 4090 cheap AF, to desolder the 12VHP and replace w/ an XT60 XT90.

 

[vinay2070]

Hats off to the connector for lasting 1 full year! You did well!

 

[vinay2070]

I sure hope so.
It'd be nice to get a 4090 cheap AF, to desolder the 12VHP and replace w/ an XT60 XT90.

I would like to know why they dont use XT90s. They are so reliable and fool proof. Used plenty in the yester years when flying RC Planes and Helis. Even the XT60s barely got warm with 50 Amps through them for more than 10 mins.

 

[InvalidError]

Even the XT60s barely got warm with 50 Amps through them for more than 10 mins.

10mis is one thing, 10 hours inside a PC case, not necessarily in a well-ventilated location is another. But yeah, I'd expect an XT60/90, EC5 or similar connector to hold up much better to abuse than MiniFitJr pins. With only one fat pin and wire each way, the risk of current balance issues is eliminated altogether.

 

[Avro Arrow]

Personally I'm not concerned. As mentioned this is statistically low, nobody else has reporter melting since the whole "you're doing it wrong statements" about plugging it in all the way.

It's a butthole design that we are stuck with. I out of fear of obsolescence bought a 600w 12VHPWR to 4x8 cable. It's inevitable that 12VHPWR connector will die or be replaced. When that happened finding power supplies with native 12VHPWR will be slim pickings, slimmer than now.

I think this whole situation is worth being aware of but not fearmongering over.

I don't think that it's about fear-mongering. To me anyway, it's more about "Why the hell did nVidia do this when the connectors we already had were working just fine?". It would be like if they suddenly decided to use some weird new USB connector with no real benefit to it and then some of them started melting and/or shorting out, damaging what they were plugged into. People would be like "WTH BRUH?" and this is no different.

My RX 7900 XTX uses three standard 8-pin PCI-Express Supplementary Power Connectors and it works perfectly. If I had paid $1600USD for a video card (which for my purposes would've meant I was fit for the Funny Farm), just the knowledge that this could happen would be another stressor in my subconscious, something that nobody needs. I would be extremely annoyed that despite having paid through the nose for my video card, I was still subject to having to deal with crap like this. Not necessarily that my card would melt, but the knowledge that it could melt.

If I'm going to accept getting fleeced, I'm damn well going to expect nothing less than complete peace of mind as a part of the transaction, not for the manufacturer to accuse me of not knowing how to properly use a power connector. This connector was obviously not ready for prime-time but nVidia forced it through anyway. To this day, nobody outside of nVidia really knows why.

 

[InvalidError]

I don't think that it's about fear-mongering. To me anyway, it's more about "Why the hell did nVidia do this when the connectors we already had were working just fine?"

Except Nvidia didn't do this alone: 12V HPWR is also a PCI-SIG spec.

While the old connectors may have worked just fine in the DIY space, needing 3-4X separate 8-pins cables isn't particularly attractive to SIs which is where the PCI-SIG gets its marching orders from.

 

[Avro Arrow]

Except Nvidia didn't do this alone: 12V HPWR is also a PCI-SIG spec.

While the old connectors may have worked just fine in the DIY space, needing 3-4X separate 8-pins cables isn't particularly attractive to SIs which is where the PCI-SIG gets its marching orders from.

I know that it's a PCI spec, but who else uses it other than nVidia?

 

[InvalidError]

I know that it's a PCI spec, but who else uses it other than nVidia?

Likely a bunch of companies the consumer space doesn't know much about for things like powering a 24-48 bays E3 backplanes.

 

[Avro Arrow]

Likely a bunch of companies the consumer space doesn't know much about for things like powering a 24-48 bays E3 backplanes.

Yeah, that sounds likely. Maybe we should see if they have similar issues. I'm guessing not because this seems like an RTX 4090-specific problem because I don't remember reading about the same problem with the RTX 4080 or 4070 Ti.

I guess the real question is... What advantage does the 12VHPWR connector offer that makes it more compelling than standard PCI-Express Supplemental Power connector? Like, I know that it can take 16ga wire but I don't remember that being an issue. Changing a standard should only occur if the system that was in place became inadequate. Since that clearly hasn't happened, I'm forced to wonder why someone thought it would be worth the hassle to use something else.

 

[helper800]

False equivalency. If power connectors of all kinds melted nearly as commonly as the 12VHPWR, the reaction to it happening wouldn't have been what it was. People would've said "Yeah, that kinda stuff happens all the time." but that's not what happened. Show me an article that details an issue surrounding standard PCI-Express Supplementary Power Connectors that warns about them melting if you don't do <insert instruction here> and then you might have a leg to stand on.

You saying he made a false equivalency, was itself a fallacy of fallacies. You said, "we'd be seeing electrical connectors of all kinds melting." You cannot say he made a false equivalency because he provided evidence that was a direct counter to your claim. You asked for evidence, and he gave you some to undercut your argument. It is not some logical mistake like a false equivalency just because you did not like the evidence provided and how it debased your argument. You seem to think that the only real evidence of the existence of other connectors melting is an article about it, which in itself is illogical. If you say something does not happen and then I provide you photographic proof it did, you cannot logically say that photo is not real evidence and that only articles count.

 

[Order 66]

Yeah, that sounds likely. Maybe we should see if they have similar issues. I'm guessing not because this seems like an RTX 4090-specific problem because I don't remember reading about the same problem with the RTX 4080 or 4070 Ti.

I guess the real question is... What advantage does the 12VHPWR connector offer that makes it more compelling than standard PCI-Express Supplemental Power connector? Like, I know that it can take 16ga wire but I don't remember that being an issue. Changing a standard should only occur if the system that was in place became inadequate. Since that clearly hasn't happened, I'm forced to wonder why someone thought it would be worth the hassle to use something else.

I agree, I think the reasoning was that eventually if enough people get ATX 3.0 PSUs with the native 12VHPWR connector was that there would be less cable clutter in general. (one 16 pin vs multiple 8 pin connectors) But there should have been further testing before the connector was used.

 

[SyCoREAPER]

I agree with everyone but everyone keeps skirting around the main focus of the issue. Focus being, it supposedly melted after a year's use. We haven't seen any other stories of connector issues.

There's no denying it's a bad design with (probably) good intentions, that's irrelevant at this point. What the focus now is why this is thiseven a news article about one reddit post.

TH should have never published this article because it's not based on any factual or even pertanent information, it's just a reddit regurgitation.

 

[instawookie]

I agree with everyone but everyone keeps skirting around the main focus of the issue. Focus being, it supposedly melted after a year's use. We haven't seen any other stories of connector issues.

There's no denying it's a bad design with (probably) good intentions, that's irrelevant at this point. What the focus now is why this is thiseven a news article about one reddit post.

TH should have never published this article because it's not based on any factual or even pertanent information, it's just a reddit regurgitation.

Hopefully this doesn't become the trend of Tomshardware, and we receive News/Articles/Data similar to what could be received from Linus Tech tips .

 

[helper800]

I agree with everyone but everyone keeps skirting around the main focus of the issue. Focus being, it supposedly melted after a year's use. We haven't seen any other stories of connector issues.

There's no denying it's a bad design with (probably) good intentions, that's irrelevant at this point. What the focus now is why this is thiseven a news article about one reddit post.

TH should have never published this article because it's not based on any factual or even pertanent information, it's just a reddit regurgitation.

I disagree. I think that letting people who still have graphics cards, 4090s in particular, know about a potential future issue is news worthy.

 

[InvalidError]

I guess the real question is... What advantage does the 12VHPWR connector offer that makes it more compelling than standard PCI-Express Supplemental Power connector?

For an SI, that would be fewer parts to spec, fewer parts to order, fewer parts to stock, fewer parts to install, cable-manage, etc.

 

[Avro Arrow]

I agree, I think the reasoning was that eventually if enough people get ATX 3.0 PSUs with the native 12VHPWR connector was that there would be less cable clutter in general. (one 16 pin vs multiple 8 pin connectors) But there should have been further testing before the connector was used.

Sure, but cable clutter is generally handled by cable sleeves. Like, people are getting so crazy with cable management that eventually everything in a PC is going to be powered wirelessly! /s

You should've seen the interior of PCs in the 90s because most were rats' nests. There was no cable sleeving nor were there modular PSUs. We also didn't really care because the idea of having a glass side panel so we could see our CPU coolers spin would've been considered laughable. I used to stick excess wires into an empty HDD bay just so that they didn't get caught up in any of the fans. Beyond that, we didn't care as long as it worked.

What's next, people are going to start using automotive wrap for their cases?

(It is not lost on me that I might be giving people bad ideas with that question...)

 

[Avro Arrow]

For an SI, that would be fewer parts to spec, fewer parts to order, fewer parts to stock, fewer parts to install, cable-manage, etc.

That's fair but it's still a Mickey-Mouse way of handling it because there will still be cables to deal with. I think that it would be more advantageous just to add this power-handling capability to the motherboard itself. Then there would be no cables at all. ASUS' "power toe" idea is actually a great idea, if something like it could be standardised under the ATX umbrella.