RTX 5090 Catches Fire on Christmas Eve — Connector Charred Beyond Melting

A festive gaming session turned into a scene of smoky alarm for a PC enthusiast on Christmas Eve, as their highly sought-after NVIDIA GeForce RTX 5090 Founders Edition reportedly suffered a catastrophic failure, with its power connector catching fire. Disturbing images shared online show the advanced 16-pin power connector, known as 12V-2×6, not merely melted but visibly charred and burned.

This alarming incident, emerging just months after the flagship GPU\’s launch, sends a chilling echo of the connector meltdown issues that plagued its predecessor, the RTX 4090. Despite a redesigned connector and stricter manufacturing specifications intended to solve the problem, this event raises urgent questions about the safety margins of next-generation power delivery under extreme loads.

From Festive Frames to Flames

According to the user\’s report on a popular hardware forum, the system was running a demanding game when a \”burning smell\” suddenly filled the room. The PC was instantly powered down. Upon inspection, the damage was severe: the plastic housing of the connector on the GPU side was blackened and deformed, with clear signs of arcing and heat so intense it went beyond melting to actual combustion. The expensive 12VHPWR cable adapter was fused to the port.

\”The smell was unmistakable—it wasn\’t just hot plastic, it was electrical fire,\” the user stated. \”This was supposed to be the fixed, safer design. I had the cable clicked in fully and wasn\’t using excessive bends.\”

A Ghost of Problems Past?

The original 12VHPWR connector on the RTX 4090 faced widespread scrutiny and user reports of melting, often linked to connectors not being fully seated or cables being bent too tightly too close to the port. NVIDIA and its partners responded with a revised \”12V-2×6\” design for the RTX 5090. Key improvements included a shorter sense pin design to ensure a secure connection is made before power flows, and a more robust physical housing.

This Christmas Eve failure suggests that while the redesign may have reduced the incidence of issues, it may not have entirely eliminated the risk. Potential factors being debated by the community and experts include:

  • Transient Power Spikes: The RTX 5090, like its predecessor, can generate massive, instantaneous power draw spikes that may exceed the nominal 600W rating, overwhelming even a properly seated connection.

  • Manufacturing Variability: Microscopic flaws in the connector or cable terminals, poor solder joints, or substandard materials in any part of the power delivery chain could create a point of high resistance and heat.

  • User Error, Redefined: While full insertion is easier to verify, the extreme physical force required to unseat the new connector might give a false sense of security, potentially masking an underlying micro-fracture or imperfection.

NVIDIA Yet to Comment; Community on High Alert

NVIDIA has not yet issued a public statement regarding this specific incident. The GPU involved is expected to undergo an RMA (Return Merchandise Authorization) process, which will likely lead to a forensic analysis by the company\’s engineers.

In the meantime, the PC hardware community is on high alert. Owners of high-end GPUs, particularly the RTX 5090, are being urged to:

  1. Visually Inspect their power connections regularly for any signs of discoloration, melting, or unusual smell.

  2. Ensure Proper Installation by using the official cable (or a certified high-quality PSU-native cable), connecting it straight into the GPU without sharp bends for at least 35mm, and confirming the connector is firmly and audibly clicked into place.

  3. Monitor Power Usage using software tools to observe if their card is experiencing extreme power transients.

A Burning Question for the Industry

This fiery failure on a holiday night is more than an isolated user\’s misfortune; it is a stark reminder of the extreme engineering challenges at the bleeding edge of performance. As GPUs push towards and beyond the 600-watt thermal design power (TDP) barrier, the industry\’s reliance on increasingly dense power connectors is being stress-tested in the most dramatic way possible. For NVIDIA, a swift and transparent investigation is not just about one customer\’s card—it\’s about maintaining confidence in the safety and reliability of its most prestigious and powerful product.