Argomenti trattati
The world of high-performance graphics cards is fraught with challenges, and the recent incident involving the RTX 5090 serves as yet another reminder of the potential pitfalls. Reports have surfaced of yet another RTX 5090 experiencing a melted 16-pin connector, adding to the growing list of hardware mishaps. Interestingly, the GPU-side connector remained unscathed, but the yellow-tipped cable—a feature intended to ensure proper seating of the 12V-2×6 cable—was severely charred. This situation has sparked discussions about the reliability and safety of these advanced GPUs.
Details of the incident
This specific case was documented by a user on Reddit, who shared that they were using an MSI Gaming Trio RTX 5090 powered by a Corsair HX1500i power supply, compliant with ATX 3.1 standards. The setup had been running for about two months when the incident occurred. Instead of utilizing the native 16-pin to 16-pin cable, the user opted for an MSI yellow-tipped 4x 8-pin to 16-pin adapter, mirroring an earlier incident that raised similar concerns.
While the user did not confirm whether the cable was fully inserted, the design of the yellow tip is meant to indicate proper seating. However, the visible burn damage along the entire row of 12V terminals raises questions about the integrity of the connection. Initially, the discoloration made it seem like part of MSI’s design, but it soon became clear that something was amiss.
The implications of design choices
Upon inspection, the GPU-side connector showed no signs of damage, which was verified by testing with a replacement cable. This incident, along with others, points to potential flaws in Nvidia’s electrical design choices for the reference boards of their RTX 40 and RTX 50 series GPUs. Previous models, like the RTX 30 series, featured implementations that included three shunt resistors. These allowed the GPU to read the six 12V pins as three separate inputs, enabling it to detect any missing connections.
In contrast, the newer designs have consolidated power distribution across all six pins to a single source. As a result, if a few pins become disconnected, the GPU lacks the ability to balance power or shut down, potentially leading to catastrophic failures. This design oversight could allow up to 500W of power to be funneled through a single pin, which is only rated for a maximum of 9.5A.
Potential solutions and community responses
One proposed solution is to use graphics cards that can provide per-pin current measurements, such as Asus’ Astral series. However, some DIY enthusiasts have taken matters into their own hands by creating custom connectors equipped with alarms and built-in fuses to prevent such failures. This innovation highlights the community’s proactive approach to addressing these significant design flaws.
As technology continues to evolve, it’s crucial for manufacturers to heed the warnings presented by incidents like this one. The community’s response and the solutions being developed serve as a testament to the resilience and determination of hardware enthusiasts. While the RTX 5090 is undoubtedly a powerhouse in the gaming world, the growing concerns regarding its reliability require attention and action from both consumers and manufacturers alike.