One Hundred RTX 4090s With Melted Power Connectors Repaired Every Month, Says Technician

12VHPWR
(Image credit: Humble-Brilliant-654/Reddit)

The saga continues over a year after the first problems with GeForce RTX 4090's melted 12VHPWR connectors were reported. A NorthridgeFix repairman claims he must fix about a hundred GeForce RTX 4090 graphics cards with failed connectors every month. He insists the connector failures are not user error, but issues with the connector design.

"We get about 20 to 25 [GeForce RTX] 4090 [graphics cards] a week," the repairman said in a NorthbridgeFix YouTube video."

This volume highlights just how common the issue is with this model. To illustrate how many GeForce RTX 4090 graphics boards come to repairs, the repairman said he had to get a half-mask respirator and an air purifier to protect his health when he fixes the melted connectors.

"We get them in so much that I bought myself this Hiroshima mask, so I do not have to smell the burn on those conductors," he said. "It smells like fireworks times ten. It cannot be healthy to keep smelling burnt connectors." 

Back in November 2022, Nvidia said that it was aware of 50 cases of melted GeForce RTX 4090 power connectors and that the standard issue was that they were not fully plugged into the graphics card, which essentially meant that it was a user error, not a failure of the connector.

Although Nvidia denies that the issues with melting power connectors on some of its GeForce RTX 40-series graphics cards lie in the peculiarities of the 12VHPWR auxiliary PCIe power connector itself, PCI SIG has now redesigned the power plug. The new 12V-2x6 addresses the main drawback of the 12VHPWR, its fixing force, and alters other aspects of the specification. This suggests that the 12VHPWR connector specification was so flawed that PCI SIG had to redesign it.

"So now it is safe to say the [GeForce RTX] 4090 melted connector is not a user problem," the repairman said. "We discussed this many times in the past, but I want to mention one last time that this is not a user error."

Edit, 11/12/2023: Corrected title.

Anton Shilov
Contributing Writer

Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.

  • peachpuff
    Time for another investigation from Igor and Steve... shirley they'll get it right this time.
    Reply
  • Hresna
    I’d like to have heard more of his reasoning for why its a design flaw and not user error.

    I’m not sure why it coudln’t be both. A connector that makes it exceedingly difficult to actually seat properly is a design flaw - but one that can be mitigated by extra caution by the user.

    If he’s only getting burnt cards and not dismantling the builds, there isn’t really any way he can know for certain these connectors were “properly seated” when the damages occurred except to take users’ word for it.
    Reply
  • YouFilthyHippo
    I have been telling people since the day this problem started that it is not user error. No one ever listens. A recall is in order, but NVidia cares too much about profits. What other computer repair technician do you know had to buy a <Mod Edit> gas mask just to do his job of repairing computer components? This is insane. He's getting 25 a week? Across a full year its been, that's gotta be 1300 GPUs JUST AT HIS SHOP ALONE. Imagine the statistics across north america, or even the world. It's time for NVidia to stop deflecting blame and own their mistake.
    Reply
  • PEnns
    Hresna said:
    I’d like to have heard more of his reasoning for why its a design flaw and not user error.

    I’m not sure why it coudln’t be both. A connector that makes it exceedingly difficult to actually seat properly is a design flaw - but one that can be mitigated by extra caution by the user.

    If he’s only getting burnt cards and not dismantling the builds, there isn’t really any way he can know for certain these connectors were “properly seated” when the damages occurred except to take users’ word for it.
    Why not call it a design flaw? I haven't heard of any other GPU that has this exact problem and of this magnitude.

    Or is Nvidia so infallible in its fans' dreams??
    Reply
  • -Fran-
    The proof is always in the puddin'.

    Regards.
    Reply
  • cyrusfox
    YouFilthyHippo said:
    I have been telling people since the day this problem started that it is not user error. No one ever listens. A recall is in order, but NVidia cares too much about profits. What other computer repair technician do you know had to buy a <Mod Edit> gas mask just to do his job of repairing computer components? This is insane. He's getting 25 a week? Across a full year its been, that's gotta be 1300 GPUs JUST AT HIS SHOP ALONE. Imagine the statistics across north america, or even the world. It's time for NVidia to stop deflecting blame and own their mistake.
    Show me one person who was refused an RMA... Until that happens enmass, there is no need for a recall. All the manufacturers will silently replace until they are out of warranty (3 years generally).
    Northridge has a reputation, even the OEM send him cards to fix... he is a refurbisher likely unwittingly... I wouldn't extrapolate his numbers to others.

    Nvidia will not need to own this as these cards sellout everywhere, there is not enough outrage. Just flame ups in the comments. Its a $2000 card(MSRP was $1600-1800 originally), maybe it will burn some users out (I for one won't ever buy another Samsung appliance for similar reasons). But overall Nvidia has zero reason to re-enter this discussion. Per MLID the 4090 is going to be phased out (very tight/nonexistent supply going forward), the coming super lines will take over. It is now understood this can happen and the mechanism whether user error or design(or a combination) has been hotly debated and its old news that has already sorted itself, new connector modification done to improve mating connector, should eliminate or lower its occurrence hopefully

    PEnns said:
    Why not call it a design flaw? I haven't heard of any other GPU that has this exact problem and of this magnitude.
    Or is Nvidia in so infallible in its fans' dreams??
    Look at power draw of 4090, some have reported it can peak draw 780W, there were reports of it blowing PSU or causing power faults, its simply a beast, both in power draw and performance (peerless). That coupled with a new power delivery standard is prone to teething issues such as this
    Reply
  • kira-faye
    Not saying it's not a problem - 100 average a month is bad, for sure - but 100 a month isn't "hundreds" a month.

    Do better.
    Reply
  • AgentBirdnest
    Headline says "hundreds" per month. One hundred is not "hundreds".

    edit: ninja'd
    Reply
  • ezst036
    Are all of these GPUs out of warranty?

    Does he upgrade the connector to prevent future melting?

    https://www.tomshardware.com/news/16-pin-power-connector-gets-a-much-needed-revision-meet-the-new-12v-2x6-connector
    https://www.tomshardware.com/news/rtx-4090-silently-upgraded-with-12v2x6-power-connector
    Reply
  • P1nky
    YouFilthyHippo said:
    I have been telling people since the day this problem started that it is not user error. No one ever listens.
    And who are you so people should listen to you? What are your credentials?
    Reply