CableMod has been having to deal with its share of brush fires following reports of its 12VHPWR adapters melting and has issued a statement regarding it. The premium custom cable and adapter company is standing behind its products and is offering to reimburse customers whose GPUs were damaged and then denied RMA request by the graphics card manufacturer. CableMod has said in its statement that it found in its investigations into the melted adapters the majority were due to them not being fully inserted but some were actually caused by manufacturing defects.
CableMod adds that it has sold tens of thousands of the adapters and it has had roughly twenty cases reported. It has also said that it is planning giveaways for cards that it has repaired and has already given away one on an NVIDIA subreddit page. Lastly, CableMod is covering shipping costs for cards that it has asked owners to send to repair shops, even if it has already been determined that improper insertion was the root cause.
From CableMod (per its Subreddit page):
“Hello Hardware Community,
Our 12VHPWR Angled Adapters have undoubtedly been our fasting-selling item at launch, and we have to start by thanking everyone who supported us and decided to pick one of them up.
Over the past couple of weeks, a number of people have been making posts regarding melted adapters, and this has caused a number of people to write to us here on Reddit and other platforms asking if this is happening a lot. We wanted to make this post to explain what we are seeing on our end, and what we are doing to handle these cases.
From the beginning, we set out to create a product that not only looked good aesthetically but was both robust and reliable. We have sold tens of thousands of these adapters since launch back in March, and of these, have received around 20 cases where the adapter connector had melted – an extremely low failure rate.
Upon analysis, the majority of failures were due to the adapters not being fully inserted into the GPU, while other cases were indeed due to defective adapters. We cannot deny that things can go wrong during the manufacturing process – no company can guarantee a 0% failure rate. However, what we can do is help our users when such things occur.
To that end, we have been closely monitoring cases that are being reported to our support team. For GPUs where the RMA requests are denied, we have offered to reimburse the full purchase price to the affected users. Some users have been asked to send their GPUs to repair centers, and we have covered all shipping costs as well. We have been doing this even in cases where the adapter was not fully inserted, as we want to make sure that none of our users are left out in the cold.
As for GPUs that are repaired, we are planning a series of giveaways to help benefit our community. We have given one away on the NVIDIA subreddit and will be giving away more across our various social platforms.
CableMod stands 100% behind each and every product that we ship, and when cases like this occur, we will always be here to help. As always, we thank everyone for their support and faith in us.
Sincerely,
Your Friends at CableMod“
External Testing
Reports of the melting adapters have been gathering for some time but the issue was put into the spotlight about a month ago when a repair shop said it had received a number of graphics cards that had been damaged from using them. Seemingly unbeknownst to the repair shop the majority of cards had actually been sent in by CableMod themselves for repair and external investigation into the damage. This detail was uncovered by Gamers Nexus after it too learned of the matter and delved further into what was happening.
Discussion (5 replies)
Join Discussion →While I'm not a fan of their tendency to magically manifest in comment form at the slightest mention of their brand (across any and all social media), this is how you treat your customers. Kudos.
Full disclosure, I own one of their cables. I just wanted to be able to close my damn case side panel and it did the trick.
Just a heads up that reports of this happening are stacking up on Reddit. It's estimated to be around 30-40 at this point and someone said that the increased reports only add 0.01 percent on top of the previous ones still making them well under acceptable levels. I doubt those with fried cards feel the same. I have noticed the melting points are often similar to the reports of when all this started happening back when the RTX 4090s came out which used manufacturer cables/adapters.
Side note: Most are probably aware that a user reported their BeQuiet cable melted at the PSU last week. This is a 1st. However, Igor's Lab did its own testing to try and replicate it and determined, like so many other hardware testing review sites, that improper plugging, or poor contact, of the cable adapter, is likely the cause.
From Igor's Lab:
"A less than ideal contact of the 12VHPWR connector on the load side (graphics card) can significantly increase its operating temperature, causing it to melt if too high loads are applied, especially at high ambient temperatures. However, a less than ideal contact of the 12VHPWR connector on the load side cannot affect the connector on the power supply side if it is properly connected!"
This has been 99% of the problem and it always goes back to this being user error. Again not the best design, but I and many others that take our time plugging in the power connector to our expensive video cards don't have an issue.
I agree. All three of my rigs are using OC'd cards (for the 3090 TI s they are usually clocking 2085 (air) and 2160 (AIO) and then 4090 2985+ (AIO)) that will pull 320-420 W during 5-8 hr gaming sessions. I've now confirmed this with Metro Exodus, Resident Evil Village/2/ and Biohazard, Witcher 3 Enhanced, Dead Space Remake, and now Hogwarts Legacy. I've done experiments withs DLSS on/off which can significantly affect power draw at 4K. I mostly game at 4K (60-120 Hz) and 5120x1440 (100 Hz).
Very rarely I've seen the 4090 pull upwards of 520+ but that was due to some excessive VRAM overclocking testing and it's extremely rare to see. Meanwhile, I'm now approaching over 1+ year with the 3090 Ti s in two rigs. I'm constantly on the alert with these, which is unfortunate, but so far so good.
I also agree that it's not the best design. Igor, and others, have said that PCI SIG is working on a new recommendation, and update for this standard regarding proper use (it's mentioned at the end of the article I linked to for Igor) but they cannot reveal more until it's officially announced. I'm happy with one connector but it would've been nice if it was beefier and had less to think about in using.
I totally agree with you on this. Hopefully it gets better in the next implementation.