Nvidia's incendiary 12VHPWR connector consumes its latest victim — Chinese RTX 3060 Ti melts down despite modest power draw after ASUS shipped it with the wrong cable

RTX 3060 Ti Melting Connector & Cable
(Image credit: Baidu)

The drive to push the performance envelope of modern GPUs often comes at the cost of power. As NVIDIA's high-end cards keep climbing in performance, so does the frequency of their power connectors melting down. This time, however, it's not even a flagship model like the 4090 or 5090, but rather a budget GPU from the Ampere series — the RTX 3060 Ti.

A user on Baidu reported the power connector of their RTX 3060 Ti GDDR6X melting down in a freak incident, specifically the Asus Megalodon V2. It's worth noting that this is a special variant of the 3060 Ti designed exclusively for the Chinese market and, as such, it features a 12VHPWR power connector instead of the standard 8 or 6-pin that was the default with the 30-series.

RTX 3060 Ti 12VHPWR connector

(Image credit: Baidu)

However, the cable that the user received was a standard 12-pin cable that lacked the additional 4 pins present in a normal 16-pin 12VHPWR connector. These extra four pins are purposefully there to ensure a secure connection between the card and the cable, precisely in order to combat situations like this one.

Therefore, this can be chalked up to a strange and unacceptable mistake on Asus' part, especially when you dig a little deeper. In the photos supplied by the user, a dual 8-pin to 12-pin adapter can be seen, which ships with some 12-pin GPUs for use with older power supplies that don't have these new connectors.

So, despite the 3060 Ti rocking a 12VHPWR connector, the user received a 12-pin cable (and adapter) in the box, both of which are catalysts for a fire hazard. But it doesn't end there. All this happened while the card does not even need the beefy 16-pin connector in the first place, as it's just a 3060 Ti, which has a maximum power draw of 310W.

RTX 3060 Ti 12-pin power cable

(Image credit: Baidu)

12VHPWR connectors don't have an issue supplying power up to 400W; it's only beyond that threshold where it starts to overheat, regardless of how tight the connection is. Here, the missing four pins that forced a 16-pin connector into a 12-pin one essentially guaranteed a loose fit that would ultimately lead to its demise.

This is the first-ever case of a relatively low-power GPU getting fried like this post the introduction of 16-pin connectors. No matter where it goes, 12VHPWR always finds itself in trouble. Sure, maybe there's some user error at play here, but shipping the wrong cable out of the box is a far more dangerous precedent to set than victim blaming.

Follow Tom's Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.

TOPICS
Hassam Nasir
Contributing Writer

Hassam Nasir is a die-hard hardware enthusiast with years of experience as a tech editor and writer, focusing on detailed CPU comparisons and general hardware news. When he’s not working, you’ll find him bending tubes for his ever-evolving custom water-loop gaming rig or benchmarking the latest CPUs and GPUs just for fun.

  • TheyStoppedit
    Umm.... the 3060Ti is 200W, not 310W. This should not happen, unless they used a 22 AWG cable that they paid 51 cents for off wish.com ...... my guess is that there was a fault in the power circuitry someone that caused it to try and pull 200W from 1 of the 6 wires. Either way, that's wierd
    Reply
  • mikeztm
    TheyStoppedit said:
    Umm.... the 3060Ti is 200W, not 310W. This should not happen, unless they used a 22 AWG cable that they paid 51 cents for off wish.com ...... my guess is that there was a fault in the power circuitry someone that caused it to try and pull 200W from 1 of the 6 wires. Either way, that's wierd

    This could happens to any card that is >120w TGP.

    Every pin should have a current of < 10a and in worse case scenario all current goes through just 1 pin that will give you 120w max for safe operation.

    Just look at the melted connector you can immediately spot only 2 pins are melted.

    Those 2 pins got more than 10a current for sure.
    Reply
  • TheyStoppedit
    mikeztm said:
    This could happens to any card that is >120w TGP.

    Every pin should have a current of < 10a and in worse case scenario all current goes through just 1 pin that will give you 120w max for safe operation.

    Just look at the melted connector you can immediately spot only 2 pins are melted.

    Those 2 pins got more than 10a current for sure.

    Then there is something wrong with the card itself. With a current of 10A, 2 wires would be 240W. The 3060Ti is only 200W, and that assumes not a single watt goes through the PCI-e slot which can take 75 itself
    Reply
  • mikeztm
    TheyStoppedit said:
    Then there is something wrong with the card itself. With a current of 10A, 2 wires would be 240W. The 3060Ti is only 200W, and that assumes not a single watt goes through the PCI-e slot which can take 75 itself
    GPU PCB need to be designed to get power from 75w PCIe slot. We can assume most GPU today takes less than 5w from the slot.

    And as I said 120w is the safe limit. This card could be shoved 200w through just 1 pin and after it’s burned current goes to the second one and burn that one in seconds.
    Reply
  • Penfolduk
    The saga around this class connector and cable standard gets more ridiculous by the month.

    You would think that, if the 4 sense pins were missing entirely, the card would just shut down without drawing any more power. Apart from whatever small power draw it needs from the PCI-E slot to run an initial check.

    Whilst I do realise the majority of cases may be down to "user error", the fact that such user error is possible is disturbing.

    I'm sure if the UK 13Amp plug and socket combination had this kind of failure rate they would never have been so widely adopted in the first place.
    Reply
  • ezst036
    This is the first time I can remember seeing this on an 30x0 card. 🤔

    It's just terrible that Nvidia doesn't give up and stop doing this to its customers. This connector has gotta go.

    These 30x0 class cards are even supposed to be properly shunted, that must have been one heck of a super cheap wire. This should just not happen.

    This should just not happen.
    Reply
  • edzieba
    Penfolduk said:
    Whilst I do realise the majority of cases may be down to "user error", the fact that such user error is possible is disturbing.
    'User error' is always possible and can never be eliminated: nothing is foolproof to a sufficiently talented fool. There are no shortage of photos of Molex Mini-Fit Jr. (AKA 6-pin and 8-pin PCIe) connectors melted out of GPUs in the same manner as Molex Micro-fit (AKA 12VHPWR), after all.

    If you want the standard changed, go to PCI-SIG who set the standard and select the physical connector.
    Reply
  • helper800
    I would bet good money that the cable was inappropriately installed causing undue resistance.
    Reply
  • aberkae
    Admin said:
    Wherever the 12VHPWR connector goes, it brings death with it. The latest in the series of murders is a measly RTX 3060 Ti from China.

    Nvidia's incendiary 12VHPWR connector consumes its latest victim — Chinese RTX 3060 Ti melts down despite modest power draw after ASUS shipped it w... : Read more
    Any 9070xt Nitros reported?
    Reply
  • helper800
    aberkae said:
    Any 9070xt Nitros reported?
    Not that I have seen or heard of.
    Reply