Sunday, September 13th 2026

ASRock's TempGuard Failed to Stop a PSU-Side Connector Meltdown on a $15,000 RTX PRO 6000 Rig

A repair shop technician has reported another melted 16-pin power connector on an ASRock Taichi TC-1650T PSU, paired with a $15,000 NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPU. The former ships with native 12V-2x6 cables carrying ASRock's own NTC temperature sensor. According to the technician's Reddit post, the connector failed catastrophically on the PSU side and melted into the socket itself, even though both ends of the cable were fully seated. The GPU reportedly survived, and the damage was confined to the power supply side. The technician also says the sensor is mounted on the wrong side of the cable, on the ground line rather than the 12 V side it's supposed to watch, which would explain why it apparently never caught the rising heat in time.

ASRock added the NTC sensor and TempGuard system to its 12V-2x6 cables specifically to shut a system down before a connector melts. This time the system failed to catch the rising temperatures, but it has worked in the past. TempGuard reportedly caught a shunt-modded RTX 5090 pulling over 1,300 W, saving its power connectors from melting. On the flip side, this isn't the first time this exact GPU and PSU pairing has run into trouble either, as a similar melting incident on the same combo surfaced back in March. If the sensor placement is really off on at least some cables, that's a bigger concern than one bad unit, especially when paired with a card that costs as much as a used car.
Source: Reddit
Add your own comment

169 Comments on ASRock's TempGuard Failed to Stop a PSU-Side Connector Meltdown on a $15,000 RTX PRO 6000 Rig

#51
A Computer Guy
user556Yup. That's why it's always the 12V pins that fail.

At this stage of things, since manufacturers are doing nothing serious, the best solution is cut the plugs off the cables and solder the cables at both ends.
But don't you still have to deal with the connector at the PSU as well?
Posted on Reply
#52
user556
A Computer GuyBut don't you still have to deal with the connector at the PSU as well?
I did say both ends.
Posted on Reply
#53
JustBenching
LittleBroOh noes! Another day, another RTX 5090 with melted connector due to faulty design user error.

And this is only a tip of a huge pyramid made of RTX GPUs that have:
- partially melted but are not causing black screens, thus users are unaware of the situation,
- melted but cases were not published or disclosed,
- melted but sold to other owner, as the price of the card multiplied (repairing connector is pretty cheap compared to getting new card).

Unfortunately, TPU does not cover all the published cases.

This connector and badly designed power delivery rail in RTX 4080/4090/5000 series is golden eggs laying duck for NorthridgeFix.
Yes, out of 50 million cards we got another one. Failure rate has risen to an astronomical 0,00001%. Woah.
Posted on Reply
#54
Hecate91
50 million cards sold with this connector? Citation needed.

If so few cards fail, then I wonder why all of these companies have to come up with things to prevent connectors from melting.
Posted on Reply
#55
BoggledBeagle
Hecate9150 million cards sold with this connector? Citation needed.

If so few cards fail, then I wonder why all of these companies have to come up with things to prevent connectors from melting.
Why do you think companies did have to come up with accessories which deal with thick AM5 heatpreaders and CPU overheating? It is a totally optional thing that just depends on an opinion of the consumers who think their CPU should not run at 95°C.

The same way some consumers have an opinion that the problem with melting cables is horrible and serious and therefore they need to buy extra equipment to solve the "problem" even if the problem really exists mainly in their head and melting connectors are comparable with all other possible GPU failures.
Posted on Reply
#56
Shrek
_roman_Every GND line in total has the summary of the currents. as stated by kirchhoff.

There are a lot of GND pins on a power supply unit on multiple connectors, not only the 12V high power one
You are right and some of the return current is probably going through the PCIe connector.
Posted on Reply
#57
user556
BoggledBeagleWhy do you think companies did have to come up with accessories which deal with thick AM5 heatpreaders and CPU overheating? It is a totally optional thing that just depends on an opinion of the consumers who think their CPU should not run at 95°C.

The same way some consumers have an opinion that the problem with melting cables is horrible and serious and therefore they need to buy extra equipment to solve the "problem" even if the problem really exists mainly in their head and melting connectors are comparable with all other possible GPU failures.
These things are melting though. And badly and expensively too. It's not just an opinion. One day one will go on fire.
Posted on Reply
#58
Hecate91
BoggledBeagleWhy do you think companies did have to come up with accessories which deal with thick AM5 heatpreaders and CPU overheating? It is a totally optional thing that just depends on an opinion of the consumers who think their CPU should not run at 95°C.

The same way some consumers have an opinion that the problem with melting cables is horrible and serious and therefore they need to buy extra equipment to solve the "problem" even if the problem really exists mainly in their head and melting connectors are comparable with all other possible GPU failures.
AM5 heatspreaders is a whole different subject, and not horrible or serious. If AM5 CPU temperatures were a serious issue we'd be hearing about more CPU's blowing up due to temps, modern CPU's running to 95C might not be ideal for highest performance however it doesn't burn the CPU.
These connectors are melting, and it is horrible and serious, especially now on GPU's being 40% over MSRP to have a connector melt with little or no warning. This issue exists and is well documented, even on 9070XT's and 5070Ti's running at 300-350W.
Posted on Reply
#59
BoggledBeagle
user556These things are melting though. And badly and expensively too. It's not just an opinion. One day one will go on fire.
Oh sure, BTW have you seen the charred motherboards caused by VRM chip explosions? And all the horrible number of peeled off RAM and GPU chips due to PCB deformation? Disaster, right?
Posted on Reply
#60
Aleksandar_038
JustBenchingYes, out of 50 million cards we got another one. Failure rate has risen to an astronomical 0,00001%. Woah.
Inventing numbers isn't going to help.
Posted on Reply
#61
BoggledBeagle
Aleksandar_038Inventing numbers isn't going to help.
Do you have any uninvented numbers? Like the REAL FAILURE RATE of the melted connector in comparison with all the other types of GPU failures?
Posted on Reply
#62
user556
BoggledBeagleOh sure, BTW have you seen the charred motherboards caused by VRM chip explosions? And all the horrible number of peeled off RAM and GPU chips due to PCB deformation? Disaster, right?
Never seen that as an article, no.
Posted on Reply
#63
Aleksandar_038
BoggledBeagleDo you have any uninvented numbers? Like the REAL FAILURE RATE of the melted connector in comparison with all the other types of GPU failures?
Real failure rate we can have when nVidia or some of AIB publicly announce it. It is not goint to happen, so...


However, we know that failure occurs in RTX 5090 cards mainly, because they can push connector well over 600W. RTX 5080 usually can not do that.

And we also know that there were no 50 millions of sold RTX 5090 cards, so...
Posted on Reply
#64
Geofrancis
If they went to 24v or 48v with 12V-2x6, its a different enough connector that it would have been safe from being mistaken for anything else and the small pins would have worked fine.

The power sense pins should have been proper data pins that could actually communicate usefull information between the power supply and GPU like throttle the gpu if the AC power browns out or power supply is overheating as well as detailed information about the power supply that it can use to set its consuption.
Posted on Reply
#65
JustBenching
Aleksandar_038Inventing numbers isn't going to help.
True. That's why im not doing that. You realise they are selling 10 or more million gpus a QUARTER? Not a year. A quarter. My estimation was much lower than that. Here, see for yourself




Hecate9150 million cards sold with this connector? Citation needed.

If so few cards fail, then I wonder why all of these companies have to come up with things to prevent connectors from melting.
Few? How many do you think have failed? I 1000x the report numbers. One freaking thousand....

Why do companies come up with unneeded products? To sell them is the answer. Delid kits, new heatspreaders, liquid metal and all kinds of crap have been around to solve "problems".
Posted on Reply
#66
Hecate91
JustBenchingTrue. That's why im not doing that. You realise they are selling 10 or more million gpus a QUARTER? Not a year. A quarter. My estimation was much lower than that. Here, see for yourself
And you realize those numbers include AI card shipments, which aren't using any consumer type power connector at all.
JustBenchingFew? How many do you think have failed? I 1000x the report numbers. One freaking thousand....

Why do companies come up with unneeded products? To sell them is the answer. Delid kits, new heatspreaders, liquid metal and all kinds of crap have been around to solve "problems".
Only Nvidia and their AIBs have failure rates, and I doubt either would ever admit how many failed unless somebody decided to sue for a defective standard.
These cables are needed because the company who enforced this standard refuses to fix it, the problem is real and a matter of time until these connectors fail on cards with high power draw, it's obvious planned obsolescence.
Posted on Reply
#67
TheGeekn°72
AthenaI see a recall in the near future...
there will be no such thing; you'd think there would have been one back when they first released the connector on RTX40
Posted on Reply
#68
BoggledBeagle
Hecate91And you realize those numbers include AI card shipments, which aren't using any consumer type power connector at all.
Perhaps you missed the graph title in large bold letters, which says "desktop". That is a consumer computer FYI.
Posted on Reply
#69
rusty caterpillar
For those who still believe in 12V fairy tales, that 12V design can carry 600 W, take a look at those pins, in comparison, side by side and try to comprehend what is the service area for each pin.



It needs just around 85 + C on 12V heat and then cold cycles to change the position of those pins and than increase the resistance to the next warm up and so on, till it melts.

An then

Posted on Reply
#70
Athena
TheGeekn°72there will be no such thing; you'd think there would have been one back when they first released the connector on RTX40
I meant for this PSU...

maybe the 6090 will get dual connectors, but I highly doubt that Nvidia would actually want to change connectors
Posted on Reply
#71
mikesg
Aleksandar_038(Replying to "Imagine using 24 or 48V with lower Amps, instead of that 12V thing... Just imagine.")

Imagine how much more expensive it would be, how much more stress to MoFSETs would be, how much more tracing it would need...

Also, nVidia, AMD and Intel could forget about improving their efficiency, instead they could keep adding moar watts until we are back at 1 kW again... We do not want that...

Finally, backward compatibility would be gone... We can not move on from ATX standard for years, let alone switch to completely different architecture...
Expense is something that 5090 buyers don't care about, and they expect to buy a new PSU at the same time, and they expect reliability for the price they pay.

In the past Nvidia needed to continuously sell GPUs to stay in business and had to turn up the clocks to sell something better than a 10% improvement, but now that their business is 90% funded by AI they can actually do the reverse. Sell a new generation every 5 years, turn down the clocks/TDP, and have a 25% improvement.
Posted on Reply
#72
rusty caterpillar
BoggledBeagleWhy do you think companies did have to come up with accessories which deal with thick AM5 heatpreaders and CPU overheating? It is a totally optional thing that just depends on an opinion of the consumers who think their CPU should not run at 95°C.

The same way some consumers have an opinion that the problem with melting cables is horrible and serious and therefore they need to buy extra equipment to solve the "problem" even if the problem really exists mainly in their head and melting connectors are comparable with all other possible GPU failures.
Comparison is off completely. CPU throttles down if is hot Connector however if is too hot it means resistance is high, so voltage drops and than GPU is requesting even more power, melting is imminent and is no throttling but acceleration of the said process.
Posted on Reply
#73
Aleksandar_038
JustBenchingTrue. That's why im not doing that. You realise they are selling 10 or more million gpus a QUARTER? Not a year. A quarter. My estimation was much lower than that. Here, see for yourself
Don't play coy... You know very well that graph shows ALL graphic cards sold.

Out of that number, there is subset of graphics with 12 HPWR connector.

Out of that second number, there is subset of graphics that actuall pull 500W or more, where connector failure can be expected... It for sure won't appear on graphics that pull 300W, no?
BoggledBeaglePerhaps you missed the graph title in large bold letters, which says "desktop". That is a consumer computer FYI.
Desktop, yes. All desktop. Including 150W graphic cards, etc...

So, we are not looking at 50 million of GPUs that are potential issue, but less than million... maybe even half a million of GPUs that could have this issue.

Given what we know now, it is realistically to expect that number of RTX 5090 surviving couple of years of usage will be very small.

So, you can twist numbers as much as you want, situation is clear. It is nVidia and PCI-SIG mistake, and it shoud be corrected ASAP. It won't be, because no-one cares about consumers anymore...
Posted on Reply
#74
BoggledBeagle
rusty caterpillarFor those who still believe in 12V fairy tales, that 12V design can carry 600 W, take a look at those pins, in comparison, side by side and try to comprehend what is the service area for each pin.



It needs just around 85 + C on 12V heat and then cold cycles to change the position of those pins and than increase the resistance to the next warm up and so on, till it melts.
What are you talking about??? What is "the service area"? Do you really believe that the pins move due to the thermal cycles? Did you just make this up or what did you base this idea on?
rusty caterpillarComparison is off completely. CPU throttles down if is hot Connector however if is too hot it means resistance is high, so voltage drops and than GPU is requesting even more power, melting is imminent and is no throttling but acceleration of the said process.
Perhaps you did not understand, I was explaining why products exist on the market - it it because a perceived need of the consumers. Not because the product is scientifically proven to be really useful.
Aleksandar_038.... Given what we know now, it is realistically to expect that number of RTX 5090 surviving couple of years of usage will be very small.

So, you can twist numbers as much as you want, situation is clear.
I am quite envious that things are so clear to you, I personally require some real data and reliable evidence, before I have those, I cannot draw any conclusion.

The cases I see reported just quite reliably prove, that the problem exists, but nobody knows, how serious this problem is, unless we have REAL RELIABLE DATA as real failure rate in comparison with other possible GPU failures.
Posted on Reply
#75
Vayra86
GrimlingAbsolutely!
Lets be clear here, the Connector is FINE! IT JUST WORKS!
All these reported failures (most of them are fake and AI-Slop of course) are only happening, because people are lacking the necessary intelligence to use the Connector correctly!
Its either user Error (very likely), or a $5.27 trillion Companies incompetence to design a Connector (practically impossible).
Matter of fact, maybe we should pay Jensen a few bucks and pray to the DLSS god rays just for providing us with this fine piece of electrical engineering. I think that way your solution gets a small blessing and you won't be part of that minuscule 0,000000001% fail rate seeing as Nvidia sells so many GPUs that it must be great whatever they do, and no matter how early their products apparently fail. They can obviously only fail so early because of user error (and a general lack of devotion to leather jackets). Also, let's consider the fact no human was capable of designing this connector; Nvidia made extensive use of AI to arrive here.

I'm actually starting to think Jensen Huang is The Emperor and Nvidia is the start of Warhammer 40.000's Imperium; all the ingredients are there! Blind devotion of its cult following, a blind hatred for the red hordes of Chaos that occasionally have corrupted GPUs coming from the Warp... and all the Xenos trying to buy those damn GPUs along the way which Humanity really needs for their data centers.

Mind blown right here
Hecate9150 million cards sold with this connector? Citation needed.

If so few cards fail, then I wonder why all of these companies have to come up with things to prevent connectors from melting.
Melting isn't failure, the product's just trying to get attached to its buyer. It just loves you so much it never wants to leave
Posted on Reply
Add your own comment
Nov 12th, 2026 19:59 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts