In a new teardown and repair video, hardware expert der8auer investigates a failed Asus RTX 4090 TUF after its owner, Aaron, experienced sudden shutdowns and GPU detection issues. The card had functioned reliably for over a year, but problems began weeks after adding a WireView Pro power monitoring device. On inspection, the system reported zero power draw on boot, although 12V voltage was present. Removal attempts revealed that the WireView had partially melted into the graphics card’s power socket. Physical inspection confirmed signs of deformation, cracks in the plastic, and bent sense pins.
Using a precision milliohm meter, der8auer measured abnormally high resistances across the 12VHPWR pins—up to 177 milliohms—compared to just over 1 milliohm on a reference RTX 5090. While the values might be distorted due to thermal damage, the discrepancy indicates poor contact and possible current imbalance. This undermines Nvidia’s prior claim that its official adapters are safe due to built-in pin bridging. As der8auer points out, bridging on the connector body does not prevent imbalanced current flow at the individual pins where thermal damage still occurs.
Connector replacement confirms card is functional
Despite significant connector damage, the card itself remained electrically sound. Der8auer desoldered the melted power socket, cleaned the PCB, and installed a new 12VHPWR connector. He also replaced the thermal pads on the memory and VRMs using Minus Pad Advance and applied a fresh thermal interface for the GPU core. After reassembly, the card powered on without issues and successfully completed an hour-long 3DMark Speedway loop at expected temperatures: 63°C core, 76°C memory, and a stable power draw via WireView.
Persistent issues with 12VHPWR safety remain unresolved
While the card was restored to full function, der8auer concluded the video by reiterating how problematic and unpredictable Nvidia’s high-power 12VHPWR ecosystem continues to be. Even with official adapters and theoretically safe designs, melting incidents still occur. A new WireView will be sent to Aaron, and der8auer hinted at further experiments to better understand the failure conditions.
Conclusion:
This latest case highlights the fragility of current high-power GPU connector standards and contradicts Nvidia’s previous reassurances. Despite design safeguards like integrated pin bridging, real-world use continues to produce failures. The lack of a fully reliable solution suggests the “next-gen” connector may still be far from safe under all conditions.
A heavy PC user since 1989, I’ve experienced every stage—8-bit, 16-bit, and 32-bit—through the early console years. The Oculus revolution pulled me into VR, leading me to open my own VR arcade. My classic education was complemented by a daily deep dive into PC hardware and software news. With over 20 years of experience in the consumer electronics industry—spanning major financial and administrative firms—I now primarily work as a freelance writer and IT consultant. Based in the UK, I live with my family.