Human Error Reportedly Caused Latest 12VHPWR Cable Melting Incident

[T0@st]

Late last month, NVIDIA claimed that 16-pin power connector issues were a thing of the past. The controversial 12VHPWR connection standard has fueled many online debates—prompting investigations from several prominent press outlets. Following NVIDIA's latest "safety" declaration—likely by coincidence—PCM Hong Kong reported another melting incident, affecting two cables and a power supply unit. The publication's hardware reviewer was recently engaged in the "full-load" testing of GeForce RTX 5090D and RTX 5080 graphics cards. Last week's evaluation session was interrupted by notable test system instabilities—upon downing tools, the PCM staffer discovered that their 1200 W PSU had given up the ghost. Additionally, two 16-pin cables had melted at both ends—initial detective work pointed to a GeForce RTX 4090 sample card being the main culprit.

VideoCardz and UNIKO's Hardware kept close tabs on PCM's next steps—online interactions, over the past weekend, spurred a re-evaluation of circumstances. According to PCM's latest update, they noticed burn marks on the GeForce RTX 4090 test unit—the two GeForce RTX-50-series cards did not exhibit any physical damage. Post-analysis, the reviewer now suspects that an SSD failure could be the root cause. They were happy to report that all involved RTX cards have survived, and that their test platform has been re-equipped with 12V-2x6 cables. An amended VideoCardz article proposes that everything came down to a simple human error.



View at TechPowerUp Main Site | Source

 

[Daven]

All of this sounds like something out of the Pat Gelsinger handbook:

Step 1: Blame the end-users. If they love a hardware company to the point of worship, they probably suffer from self-loathing anyway.
Step 2: If Step 1 doesn't work, blame the motherboard manufacturers. No one cares about them.
Step 3: If Steps 1 & 2 don't work, offer extended warranties and replacement units but deny 99.9% of the claims on the basis that A) it's the end-users fault and B) it's the motherboard manufacturer's fault.

 

[Visible Noise]

All of this sounds like something out of the Pat Gelsinger handbook:

Step 1: Blame the end-users. If they love a hardware company to the point of worship, they probably suffer from self-loathing anyway.
Step 2: If Step 1 doesn't work, blame the motherboard manufacturers. No one cares about them.
Step 3: If Steps 1 & 2 don't work, offer extended warranties and replacement units but deny 99.9% of the claims on the basis that A) it's the end-users fault and B) it's the motherboard manufacturer's fault.

Just ignore the facts presented and steadfastly refuse to reconsider your position.

 

[bonehead123]

All of this sounds like something out of the Pat Gelsinger handbook:

Step 1: Blame the end-users. If they love a hardware company to the point of worship, they probably suffer from self-loathing anyway.
Step 2: If Step 1 doesn't work, blame the motherboard manufacturers. No one cares about them.
Step 3: If Steps 1 & 2 don't work, offer extended warranties and replacement units but deny 99.9% of the claims on the basis that A) it's the end-users fault and B) it's the motherboard manufacturer's fault.

"And now you understand. Anything goes wrong, anything at all... your fault, my fault, nobody's fault... it won't matter - I'm gonna burn your ass off. No matter what else happens, no matter who gets burned, I'm gonna burn your ass off" - Jacob McCandles (well, sorta)

 

[Legacy-ZA]

I mean, what did the guy expect? Trying to power 2x 600W GPU's with a 1000W PSU.

 

[KillSwitx]

This cable is the worst thing ever ... even the cable section/diameter is worst than using 2x8 pins ... while requesting more power...

I hope that one day someone will look back and say "yeah, lets go back to the old solution or do a new one with proper application ... "

 

[Markosz]

Human error yea... the human who designed this connector.

 

[KillSwitx]

It was Intel that presented this new connector pattern ... now how this pass all the tests and homologations its something weird, again, using wires that have lower section than the old 2x8 ... And a connector that is flawed to begin with ... can't understand...

 

[phanbuey]

You're holding it wrong!

 

[londiste]

This cable is the worst thing ever ... even the cable section/diameter is worst than using 2x8 pins ... while requesting more power...

8-pin connector has 3 12V pairs, 2x8 has 6 12V pairs in total, same as 12VHPWR/12V-2x6. While in practice manufacturers tend to use the same wire gauge for both, 8-pin mandates a thinner minimum gauge.

 

[freeagent]

I broke the clip off of my 12v2x6 and I am still using it.. works fine

 

[KillSwitx]

The 12V-2x6 is 3mm and is rated at 8.5A per terminal in a 2x6 configuration. The Mini-Fit HCS is 4.2mm and is rated at 10A per terminal in a 2x4 configuration.

8-pin connector has 3 12V pairs, 2x8 has 6 12V pairs in total, same as 12VHPWR/12V-2x6. While in practice manufacturers tend to use the same wire gauge for both, 8-pin mandates a thinner minimum gauge.

 

[Assimilator]

Wow, amazing how most of the anti-NVIDIA crew are dead silent now.

Just ignore the facts presented and steadfastly refuse to reconsider your position.

Just put the ignorant on Ignore. Your mental health will thank you.

 

[Darmok N Jalad]

Post-analysis, the reviewer now suspects that an SSD failure could be the root cause.

This just seems odd to me. I get both are connected to the PSU, but how does it take out the cable to the GPU? Seems like it could be the other way around, where the GPU is so busy pulling amps that a malfunction accidentally cooks the SSD. The GPUs are undamaged because they are designed to handle crazy power spikes, where the SSD is not.

 

[Vayra86]

What if all this along with that hardware failure thing is a big trick to play the Chinese out of their sanctioned GPUs?

Just playing Stratego here

 

[Clawedge]

 

[Hecate91]

Nvidia blamed the users, re-designed the connector, insisted it was fine and blamed the users again, and the connector still melts with very little room for any manufacturing tolerances. It's hilarious how anyone still defends a multi trillion dollar company which refuses to accept any blame for such issues.

 

[eidairaman1]

I mean, what did the guy expect? Trying to power 2x 600W GPU's with a 1000W PSU.

Here's the point, current draw with gpus is getting out of control

 

[Heiro78]

I mean, what did the guy expect? Trying to power 2x 600W GPU's with a 1000W PSU.

The TPU write up talks about a 1200W PSU and didn't indicate 2 cards were being used at the same time

 

[john_]

So, now we even consider hardware reviewers incompetent enough to not know how to insert a power cable on a video card.
Anything, just anything to go with the narrative that will help Nvidia to get away with it. (that last remark has nothing to do with the news article here)

 

[_roman_]

Has anyone really read the text from the topic post?

Post-analysis, the reviewer now suspects that an SSD failure could be the route cause.

route cause

Or was it the root cause?

Root cause analysis - Wikipedia

en.wikipedia.org

Can someone explain me why a SSD should cause this? Which formfactor of SSD? 2.5"? M2-nvme? usb-a or USB-C ?

--

Another day - another bullshit about failed NVIDIA connector. Or maybe not failed NVIDIA connector. Yesterday AMD graphic card this - AMD graphic card that. Today nvidia connector this - nvidia connector that.

They were happy to report that all involved RTX cards have survived

Everything all right - just some bad cables - for whatever reason - graphic card okay - just 50 € (fill in the currency amount you like) spend on new cables.

Nvidia buyers will say it's the user fault. AMD buyers will most likely say it's the connector fault. I doubt anyone will change their opinion by now about this connector.

 

[phanbuey]

Has anyone really read the text from the topic post?



route cause

Or was it the root cause?

Root cause analysis - Wikipedia

en.wikipedia.org

Can someone explain me why a SSD should cause this? Which formfactor of SSD? 2.5"? M2-nvme? usb-a or USB-C ?

--

Another day - another bullshit about failed NVIDIA connector. Or maybe not failed NVIDIA connector. Yesterday AMD graphic card this - AMD graphic card that. Today nvidia connector this - nvidia connector that.



Everything all right - just some bad cables - for whatever reason - graphic card okay - just 50 € (fill in the currency amount you like) spend on new cables.

Nvidia buyers will say it's the user fault. AMD buyers will most likely say it's the connector fault. I doubt anyone will change their opinion by now about this connector.

That does indeed sound like a route cause. Which is similar to a root cause, but where the object in question is totally unrelated and the blame is "routed" to it as the cause.

 

[Ols-Hol]

I don't think they suspect SSD issue as cause of the cable melting, GPU clearly is.

They thought their system was wonky due to SSD before they found about about the cable, am I right?

Also I don't see how this "turns out to be an user error" if the cable did get melted, fact. Or is it an user error to get an Nvidia card?

 

[user556]

The human error was to keep using a power supply that had a 12VHPWR plug on it. It's the cable plugs that are the problem, not the board receptacles.

All 12VHPWR cables should be returned and replaced with 12V-2x6 cables, free of charge.

And it's very poor industry practice that this hasn't been a general recall already.

 

[KillSwitx]

The human error was to keep using a power supply that had a 12VHPWR plug on it. It's the cable plugs that are the problem, not the board receptacles.

All 12VHPWR cables should be returned and replaced with 12V-2x6 cables, free of charge.

And it's very poor industry practice that this hasn't been a general recall already.

you know that 12v-2x6 and 12VHPWR cables are pretty much the same right? what changed was on the GPU connector ( female connectors ) with longer pins by 0.25mm and shorter sense pins by 1.5mm

The cables are not the issue... something like this