RTX 4000 series burning cables thread

[Batou1986]

Except they were completely fabricated. That's the issue. You can't claim to know anything when half your sources are outright fabricating data.

So what part was fabricated WCCFTECH showing connector heating up or the rebuttal that it was user error by bending the cables, because the "made up" slides from PCI-SIG where posted as a response to WCCFTECHl's testing showing the connector heating up and having inconsistent voltages?

 

[ir_cow]

@The red spirit - This can't be that hard to look up the amperage. 55A = 660 Watts. Still over what the GPU is allowed. Remember PCI-SIG really underrates these cables. Look at the 6Pin PEG. 75-watts is for those horrible 22awg cables.

 

[wolf]

That's the type of comment that helps no one. ......... and yet someone eventually claims "i had no issues".

Except it seems like people having issues are the exception to the rule, as oppose to an overwhelming majority or flood of issues and one person saying I had no issues. Seems more like your comment helped no one.

but you won't find that discussed here anongst the screeching about adapters.

Need to find something to rage about, especially if Nvidia did it!

I just like that all the names in this thread bashing the cards have AMD GPU's. Seems peculiar and like some of you have way too much free time on your hands.

Welcome to TPU

 

[80-watt Hamster]

Except it seems like people having issues are the exception to the rule, as oppose to an overwhelming majority or flood of issues and one person saying I had no issues. Seems more like your comment helped no one.

Acknowledging that this is all probably overblown, it still really shouldn't be happening at all.

 

[The red spirit]

@The red spirit - This can't be that hard to look up the amperage. 55A = 660 Watts. Still over what the GPU is allowed. Remember PCI-SIG really underrates these cables. Look at the 6Pin PEG. 75-watts is for those horrible 22awg cables.

View attachment 267010

I think that they just substracted current slew rate from amperage, not to mention allowable voltage differences. I don't see how PCI-SIG is being too cautious.

 

[R-T-B]

because the "made up" slides from PCI-SIG where posted as a response to WCCFTECHl's testing showing the connector heating up and having inconsistent voltages?

Jonnyguru who is an industry veteran on the same list basically confirmed the WCCD provided slides were fabricated yes.

It really is the blind leading the blind in here. I'm out.

 

[The red spirit]

Jonnyguru who is an industry veteran on the same list basically confirmed the slides were fabricated yes.

But do we have any reason to trust him? I'm pretty sure that he's talented, but let's not ignore the fact that he's affiliated with Corsair, one of the OEMs who don't want to deal with burned shit.

 

[R-T-B]

But do we have any reason to trust him?

one of the OEMs who don't want to deal with burned shit.

Put two and two together. Did you seriously just answer your own question?

And yes he's one of the most trustworthy established names in the PSU industry.

 

[The red spirit]

Put two and two together. Did you seriously just answer your own question?

And yes he's one of the most trustworthy established names in the PSU industry.

And yeah, he's gonna deny problems too...

 

[Dr. Dro]

Conclusion: RTX4090 (at least) needs an ATX3.0 PSU and its not about the max power but about the communication between GPU and PSU.
Adaptors with an ATX2.0 PSU are potentially unsafe

This is conjecture, all renowned PSU experts have already made their case. ATX 3.0 has no bearing on this (potential) mechanical failure, hence, having an ATX 3.0 PSU will not change anything.

I mean,

If you have the money to buy a 4090GTX, u must have the money to buy a new PSU to go with it.

saving on PSU is not the play

Anyone who owns a high-quality power supply does not need to replace theirs, especially at the ultra high end. No PSUs that advertise 3.0 support have even come close to outperforming models like the Corsair AXi series.

So what part was fabricated WCCFTECH showing connector heating up or the rebuttal that it was user error by bending the cables, because the "made up" slides from PCI-SIG where posted as a response to WCCFTECHl's testing showing the connector heating up and having inconsistent voltages?

The problem is taking WCCF seriously to begin with.

And yes he's one of the most trustworthy established names in the PSU industry.

and I'll side with whatever Jon and Aris have to say, every time.

 

[ir_cow]

I think that they just substracted current slew rate from amperage, not to mention allowable voltage differences. I don't see how PCI-SIG is being too cautious.

Okay, say those are precisely right, It is still 660 Watts, doesn't have anything to do with ATX 3.0 PSU like your claiming and the cable shouldn't be bent over. Don't forget these people will never admit they bent those cables to make it fit nicely for wire management. Conclusion. The technical specs are sound, its how companies impalement it and consumers use it.

I think the plug is ok because it must have been tested in a lab over & over again. I think the problem is the built material used of the conducting pins/plastic housing or both. Companys are known for cutting corners & this is where I think the issue is.

Delshay is spot on here.

 

[awesomesauce]

This is conjecture, all renowned PSU experts have already made their case. ATX 3.0 has no bearing on this (potential) mechanical failure, hence, having an ATX 3.0 PSU will not change anything.



Anyone who owns a high-quality power supply does not need to replace theirs, especially at the ultra high end. No PSUs that advertise 3.0 support have even come close to outperforming models like the Corsair AXi series.



The problem is taking WCCF seriously to begin with.



and I'll side with whatever Jon and Aris have to say, every time.

Sorry but i wouldn’t put adapters on a 4090. Just not a fan

and i understand on paper it work but i would prefer direct power and not adapter

but that just me

 

[Dr. Dro]

Sorry but i wouldn’t put adapters on a 4090. Just not a fan

and i understand on paper it work but i would prefer direct power and not adapter

but that just me

The adapters are a solution for power supplies that were not yet updated with native cables. I'm confident that they are safe, these aren't the same as the old Molex adapter cables.

However, I am also not a fan because I take pride in the small details and cable management, I've taken this into account delaying my upgrade for a year and opting to instead renovate my setup after an incident that occurred a while ago. EVGA may yet release 12VHPWR cables for the G2/P2/T2 series, or maybe a device like their PowerLink adapter that does the conversion going forward and by the time the RTX 4090 Ti and RX 7950 XT come out, all issues arising from the use of this new cable format should be resolved.

 

[Ando]

Put two and two together. Did you seriously just answer your own question?

And yes he's one of the most trustworthy established names in the PSU industry.

I’m not saying he’s incompetent, I own a corsair PSU he probably helped make.
But not a good look that he recently made a claim that Gamers Nexus and Guru 3D were mistaken about the very power cable being discussed in this thread. And in short order he was proven to be wrong.
It’s accurate to say we’d need real data to make a firm conclusion on the safety of the adapter (which you can get from mass usage and taking a look at the failure rate). However, you shouldn’t just “trust the experts”. They are not infallible, they can be wrong.

 

[Crackong]

I am kinda surprised this one isn't on the main TPU news yet

 

[Vayra86]

Can you cite this?


It does for most users and many reviewers.


You guys are all over the place and confusing pcie 5 connector with ATX 3.0 amongst other things.


And the only way that's fair is if you do the same crappy idea with the old pcie connectors as well.

I'm reading your typical M.O. here, there is never a problem for most users but eventually, at the end, we always conclude there was definitely something going on there. Remember crypto? And I can think of a few other examples. You're like a living 'this is fine' meme by now. But the fact is, some cards are not surviving their early days. Again. Space Invaders last gen were also quite true. Was a bad batch, but they definitely existed. EVGA FTW's burning memory to a crisp? Definitely happened. Etc etc etc. The track record is expanding rapidly the last 5 years imho.

Its clear these connectors are too weak for normal use. You're supposed to have some leeway as a user to be silly with your hardware. Some. Here, there is zero and it is crystal clear, but that was already the case when these tiny connectors got released, this is a weak point in the whole design update for Nvidia cards and its not even the first gen with them. They should have damn well known better. Why are adapters even allowed?

Heck I was already worried about ye olde 8 pins on some models. There's a lot of pull on those cables. Here we have the mini-me version that pulls higher wattage, gosh what could possibly go wrong?

Acknowledging that this is all probably overblown, it still really shouldn't be happening at all.

This. Exactly.

 

[JustBenching]

That's the type of comment that helps no one. For any issue, major or not, you always get someone that never got it. Major safety flaws on cars that can kill you, the cars get recalled, and yet someone eventually claims "i had no issues".

Ιt works both ways. From every non issue, you get someone that god it. You know how many threads there are on reddit about traditional 8pins catching on fire? Lots and lots

 

[Zach_01]

This is conjecture, all renowned PSU experts have already made their case. ATX 3.0 has no bearing on this (potential) mechanical failure, hence, having an ATX 3.0 PSU will not change anything.

I'm glad that this sentence you cherry picked wasn't the only one I wrote...

I did mention about bending, and connector quality.

Also without the 4 pin sense part there can be a potential damage on the PSU and of course I'm not talking about the Corsair AXi line which is among the very top PSUs today.

Thing is that nVidia rushed the connector type out (without the proper 100% support for it) for this kind of wattage/current and hurting this otherwise fine product.

PC component design is(?) suppose to be fool proof and/to minimize any potential damage.

You know how many threads there are on reddit about traditional 8pins catching on fire? Lots and lots

Do these lots of 8-pins caught on fire or melt within the first couple of weeks of introduction? (on their max or close to max of specification?)

 

[BSim500]

It certainly doesn't look good though when nVidia mods start panic deleting posts on Reddit talking about how absurd it looks to run with the case side panel off to adhere to a "proper spec" of 'no bend in the first 35mm of cable' in a normal sized case. If there genuinely was no real issue, they could certainly respond in a more positive way than "Quick, let's hide all talk about this in the most suspicious way possible!"

 

[JustBenching]

Do these lots of 8-pins caught on fire or melt within the first couple of weeks of introduction? (on their max or close to max of specification?)

They melted during the first week of introduction of 3080s and 3090s, possibly radeons as well but didn't google about it.

 

[MarsM4N]

The answer is simple. The connector was lose or not seated correctly and the pins were arcing.

Aside from toasted connectors, how does "arcing" affect the card itself? A unstable power delivery will have an impact on FPS, and/or might even be unhealthy for the hardware itself.

I just like that all the names in this thread bashing the cards have AMD GPU's. Seems peculiar and like some of you have way too much free time on your hands.

It's called "Schadenfreude". A natural human emotion (non exclusive for Team AMD members).

Spoiler: Schadenfreude

 

[The red spirit]

I am kinda surprised this one isn't on the main TPU news yet

I'm not surprised at all. Any legitimate journalist, before publishing news, should first verify information. What I post here simply can't be verified. I'm here more like "hey something is happening" and seeing how it ends up being in the end. I would rather call what I post as awareness raising, but not exactly news.

Okay, say those are precisely right, It is still 660 Watts, doesn't have anything to do with ATX 3.0 PSU like your claiming and the cable shouldn't be bent over. Don't forget these people will never admit they bent those cables to make it fit nicely for wire management. Conclusion. The technical specs are sound, its how companies impalement it and consumers use it.

Now tell me one thing, if 4090 is so damn wide that it doesn't fit in 95% of cases and that cable requires additional 35mm, that eliminated majority of those 5%, what else did you expect to happen? That people won't bent (even if slightly) those cables? That's not how it works and also, it was exactly the PCI-SIG themselves, who warned OEMs that adapters may not meet their own spec, therefore whoever made adapters is guilty. Regardless of any of this, in 12VHWRP mode, RX 4090 has some spikes above 660 watts, which is outside of specification anyway.

 

[Bomby569]

Except it seems like people having issues are the exception to the rule, as oppose to an overwhelming majority or flood of issues and one person saying I had no issues. Seems more like your comment helped no one.

This is less then 2 week old thing.

 

[wolf]

This is less then 2 week old thing.

Then how about we give it some time to see it addressed hey, my money is on the adapters themselves, but time will tell all, and then we can safely critisise whoever is responsible for it happening in the first place.

 

[The red spirit]

Finally more testing:

Unfortunately test was done without that infamous adapter that melted, but anyway, card seems to be fine under Furmark. At least in that test, bent connector basically showed no big difference. But yeah, it's a 12VHWRP native power supply too.

Here's another source of information:
https://www.reddit.com/r/nvidia/comments/ycwr1y
Much less legitimate and yeah, power usage was most likely measured via software, which is not terribly accurate, anyway, test was done with that adapter and during Superposition run, cable averaged at around 50C. Not bad, but as far as I know Superposition is "lighter" on power consumption of card. Anyway, nothing alarming happened during stress test, so that's something.