RTX Pro 6000 power connector melts despite ASRock’s TempGuard thermal protection feature
Table of Contents
The 16-pin power connector continues to cause problems for high-end GPUs, and this time, even a safety feature designed specifically to prevent overheating reportedly wasn’t enough. A repair shop technician has shared images of a melted connector from a system running Nvidia’s extremely expensive RTX PRO 6000 Blackwell Workstation Edition.
According to the technician, the GPU was paired with an ASRock Taichi TC-1650T, a high-end 1650W power supply that supports ATX 3.1 and PCIe 5.1. More importantly, its 12V-2×6 cable comes with ASRock’s NTC temperature sensor and TempGuard protection, which are supposed to help prevent exactly this type of damage.
Despite those protections, the connector reportedly became hot enough to melt into the socket on the PSU side. ASRock’s NTC sensor monitors the connector for excessive heat. The protection system is designed to shut the PC down if temperatures rise beyond 105°C, theoretically stopping the connector from reaching the point where serious damage can occur. That apparently didn’t happen in this case.
Latest tech deals.
- AMD Ryzen™ 9 9900X 12-Core - was $499 now $328
- CORSAIR 4000D PC Case - was $124.99 now $75.99
- Hisense 65" Class E6SR - was $549 now $377.99
- Apple 2026 MacBook Air - was $1,999 now $1,799
- Skytech Gaming PC Desktop (9850X3d) - was $4,299 now $3,799
- Apple Magic Mouse - was $79 now $59
- MSI Prestige 16 Flip - was $2,149 now $1,899
- Kotin 9700X Gaming PC - was $2,399 now $2,199
Prices correct as of October 9th, 2026.
The technician claims that the 16-pin cable was completely inserted at both ends, ruling out an improperly seated connector as the obvious explanation. However, they noticed something unusual about the placement of the NTC sensor.
According to them, the sensor is positioned on the ground side of the cable rather than the 12V side. If that’s accurate, it could explain why the system didn’t detect the dangerous temperature increase quickly enough to shut itself down.
Thankfully, the RTX PRO 6000 itself reportedly survived the incident, with the visible damage limited to the PSU side. That’s particularly fortunate considering the workstation GPU currently costs around $15,000 or even more depending on availability.
For now, it’s unclear whether this was caused by a defective cable, incorrect sensor placement, or another issue entirely. Still, seeing a 12V-2×6 connector melt despite having dedicated temperature monitoring shows that the problem hasn’t completely disappeared, even when additional safeguards are in place.
About the Author