Native 12V-2×6 cable melts on RTX 5090 system with ASRock TempGuard
We received some interesting information from one of the users affected by the melting RTX 5090 power cable. The owner, who identifies as Riptide, says the cable was fully seated at both ends and had no sharp bends. The PC operated normally for several weeks before it began showing software instability and crashes. User inspected the system and discovered visible damage to the power connector.
An ASRock PG1000-PSF power supply reportedly continued operating after its native 12V-2×6 connector began melting. The system used an MSI GeForce RTX 5090 GAMING TRIO OC and ASRock’s TempGuard cable, which is designed to monitor connector temperatures and shut down the power supply when overheating is detected.
TempGuard protection did not shut down the system
According to Riptide, the TempGuard protection n ever triggered and the computer did not shut down automatically. User stopped using the system only after finding the burned connector. The power supply was later returned to Amazon, while the graphics card was given to a friend for testing and appears to remain operational.
Source: Riptide
Worst part, this was reportedly the second RTX 5090 connector failure experienced by the same user. The previous incident involved an RTX 5090 Founders Edition and a Corsair SF1000 power supply. Riptide says both connectors were fully seated and that both failures caused more damage at the power supply side than at the graphics card. Hard to imagine one user experiencing the same issue twice on two different 5090 models.
der8auer points to uneven current distribution
The photos were shared with VideoCardz through Roman “der8auer” Hartung. Based on the visible damage, der8auer believes one terminal carried much more current than the others. He pointed to Pin 1, which appears to have generated enough heat to damage both the cable and the PSU-side connector.
Source: Riptide
The case involved a native 12V-2×6 connection between the ASRock power supply and the graphics card, without NVIDIA’s adapter. The exact cause cannot be confirmed because the power supply has already been returned, but the system reportedly continued running despite the thermal protection designed to prevent this type of damage.
Many thanks to Riptide and Der8auer for the photos!
