RTX 4000 series burning cables thread

[awesomesauce]

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

 

[DemonicRyzen666]

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

I agree, I had my cpu pin plug end up going in wrong on my old overclocked phenom II 1090T system. What happened after than is the current draw from the missing lead. ended up going over the safe maxuim on the other wirses for the cpu pin. Which lead to three of them smelling like they were melting, & ended up burning.

 

[ir_cow]

It does, because due to using adapter, cards have limited TDP and already melt connector, more TDP and even more melted connectors. Just take a look yourself:

NVIDIA RTX 4090: 450 W vs 600 W 12VHPWR - Is there any notable performance difference?

NVIDIA’s GeForce RTX 4090 requires between 400 and 600 W of power. We pair it with a Seasonic VERTEX GX-1200 to determine whether there is any difference in performance or stability between 450 W and 600 W 12VHPWR connectors.

www.techpowerup.com

Wattage peaks will be even worse than they are now, despite all the fancy marketing saying you that connector is "smart" and will communicate with PSU. TPU's review clearly shows that it's running almost 100 watts over spec with "smart" connector. I mean, it's just laughtable when literal butt plug from porn shop could handle more watts than this "smart" connector. Seriously, nV should release Hard Leather edition with really hard and chonky plugs.

How does this "power spike" to 600 watts change when a ATX 3.0 PUS is used? Doing some basic math says 6x 12v pins rated for 10A (at least) can handle 720 watts. The card at stock is 450 and maximum of 600. Still below spec. PSU isn't the problem here, or the card. Its the connector being incorrected used.

"The results of today's article lead to the conclusion that an ATX v3.0 ready PSU with a native 12VHPWR connector, is not required to operate an RTX 4090 properly. Even with the Powenetics system, with more than 1000 Hz polling rates or less than 1 ms between power measurements, I couldn't "catch" a power spike exceeding 700 W, while the PCIe 5.0 standard speaks for up to 1800 W power spikes. I cannot know what the upcoming GPU generations will bring. Still, you can safely use proper adapters and your existing PSUs for the moment, as long as they have enough max power to support the Ada Lovelace GPU's energy requirements."

Here is the conclusion if you missed it.

 

[Bomby569]

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

what magic PSU is that? the new ATX 3.0 use the exact same connector that melted and spec'd cable. The problem is not the old 8 pin cables for sure.

 

[phill]

I feel like that Lesley Nelson "Nothing to see here, please disperse" meme should be put in this thread....

In the videos I've seen with these plugs, they didn't looks the best but why did they make the connectors so small?? If its going to be a weak point on the card for any reason, surely you'd think to re-design the plug at least wouldn't you??

 

[The red spirit]

How does this "power spike" to 600 watts change when a ATX 3.0 PUS is used? Doing some basic math says 6x 12v pins rated for 10A (at least) can handle 720 watts. The card at stock is 450 and maximum of 600. Still below spec. PSU isn't the problem here, or the card. Its the connector being incorrected used.

Here's a thing, it can't handle what you say it does.

Here is the conclusion if you missed it.

Is there any basis why would those connectors tolerate 1800 watts, even if for one spike? Aren't you American? Don't you have 120 volts at home. Tell me, from where those watts are supposed to come from?

 

[dj-electric]

So far no issues with the connector, or any kind of hotspots looking through a thermal camera.
That said, not a fan of the adapter solution or the adoption of the connector without having power supplies that support native connection widespread.

 

[Bomby569]

That said, not a fan of the adapter solution or the adoption of the connector without having power supplies that support native connection widespread.

Having a ATX 3.0 would change nothing, the cable didn't melt. The connector that is problematic is the same, you still can't bend them or they risk melting.

 

[HTC]

I get you but I doubt that AMD are going to do anything different as their cards are already being manufactured as we speak......

You're assuming they didn't design their cards with different orientations for the power cables from the beginning.

Dunno if they did, or didn't: we'll have to wait until they are launched, unless there have been leaks i'm not aware of.

 

[dj-electric]

Having a ATX 3.0 would change nothing, the cable didn't melt. The connector that is problematic is the same, you still can't bend them or they risk melting.

Having so many leads connect and merge is a big part of the flexibility and hotspot issue of the connector. Having 1:1 leads will slightly ease the contact situation with its increased flexibility

 

[damric]

The solution here is we need someone willing to perform an experiment on one of these cards where they start out with straight cable insertion and slowly start bending, while pulling 600w...for science.

 

[The red spirit]

The solution here is we need someone willing to perform an experiment on one of these cards where they start out with straight cable insertion and slowly start bending, while pulling 600w...for science.

@W1zzard ?

 

[R-T-B]

The official specification calls for 35mm of straight cable from the connector base

Can you cite this?

Here's a thing, it can't handle what you say it does.

It does for most users and many reviewers.

Having a ATX 3.0 would change nothing, the cable didn't melt. The connector that is problematic is the same, you still can't bend them or they risk melting.

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

The solution here is we need someone willing to perform an experiment on one of these cards where they start out with straight cable insertion and slowly start bending, while pulling 600w...for science.

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

 

[ir_cow]

Here's a thing, it can't handle what you say it does.

Have you read the official PCI-SIG documentation? You can do the math too for this 12VHPWR cable.

 

[Solaris17]

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

I also know how to arc weld

 

[delshay]

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

Then the simple answer to the problem is to build a slightly tighter tolerance of the mating pins or make the female connector pin stronger so it not so much acceptable to opening up when cable is bent, paying particular attention to the plastic housing around the female connector.

 

[ShiBDiB]

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.

 

[ir_cow]

PCI-SIG did warn this could be problem back in early September. Something about incorrect use of the cable

 

[evernessince]

Can you cite this?

This video goes over it using two sources:

(links to those sources are also in the description of said video)

 

[The red spirit]

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.

Ironically my most used graphics "card" is actually Mali G72 MP3, also I have been bashing AMD cards a lot in the past.

It does for most users and many reviewers.

It's like saying that most users enjoy it, but only 5% of their houses burn to ruble. Anyway, I would love to see a proper investigation with thermal camera, measured temperatures and etc.

 

[AsRock]

running three 8 Pins slightly "out of spec" would have been much safer than this garbage.

And guess who's not using this dumb plug ?.

Intel, which i believe was part of the reasons it came about.

 

[delshay]

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.

 

[The red spirit]

Have you read the official PCI-SIG documentation? You can do the math too for this 12VHPWR cable.

No, because it is not public information and to view their documents you have to belong to some company. But either way:

PCI-SIG Warns of Potential Overcurrent/Overpower Risk With 12VHPWR Connectors Using Non-ATX 3.0 PSU & Gen 5 Adapter Plugs

With ATX 3.0 PSU's coming soon, there has been talk about a major safety concern affecting the 12VHPWR Gen 5 connector as per PCI-SIG.

wccftech.com

Connector was designed for 600 watts.

 

[The red spirit]

found another one.
View attachment 267008

It's already posted on second page, but thanks anyway.

 

[R-T-B]

It's like saying that most users enjoy it, but only 5% of their houses burn to ruble.

No it's like directly contradicting your statement that it just can't carry the current.