Nvidia’s recent graphics driver rollout is turning into a headache for many PC gamers and GPU enthusiasts, especially those using high-end GeForce RTX cards. What started as a “Game Ready” update meant to improve performance for Resident Evil Requiem has quickly become a chain of frustrating problems—performance drops, hardware monitoring quirks, and now a fresh controversy involving GPU voltage behavior.
After the problematic 595.59 driver reportedly caused major performance regression and even triggered GPU fan detection issues for some users, Nvidia pulled it back and pushed out a newer release to stabilize things. But the follow-up driver, version 595.71, appears to introduce another unexpected change: a voltage cap (or voltage lock) affecting certain RTX GPUs.
Based on testing shared by the YouTube channel Bang4BuckPC Gamer, at least the GeForce RTX 5090 is impacted. Other user reports suggest the issue may also extend to the RTX 4090, RTX 4080 Super, and RTX 4070 Ti Super, though the full list of affected graphics cards still isn’t fully confirmed.
In the example shown with an Asus TUF Gaming GeForce RTX 5090 OC Edition, the card reportedly hit around 1.055V to 1.060V at stock settings. The strange part shows up when overclocking: with a +150 MHz core offset, the core clock increased as expected, but the voltage dropped to roughly 1.045V to 1.050V instead of rising or remaining steady. Pushing further to a +300 MHz overclock reportedly forced voltage down even more—dropping under 1.0V (around 0.990V)—while the GPU still wouldn’t hold core clocks beyond the 3,000 MHz mark. In other words, the driver seems to be limiting voltage in a way that also caps frequency scaling, reducing the benefits of overclocking and potentially lowering peak performance.
Additional testing on an MSI GeForce RTX 5090 SUPRIM X showed similar behavior. With an older driver and manual overclocking applied, the GPU could reportedly run around 1.020V to 1.030V and boost up to approximately 3,030 MHz. After updating to 595.71 using the same overclock settings, the card reportedly wouldn’t exceed 3,000 MHz and voltage dropped to about 1.005V to 1.010V. Even at stock settings, the newer driver appeared to run the GPU at lower voltage than before.
So why would Nvidia do this? The most likely explanation is safety. Over the past months, reports of melted 16-pin power connectors and related power delivery concerns have put extra pressure on GPU vendors to reduce risk. A driver-level voltage limitation could, in theory, help lower the electrical and thermal strain passing through the connector by restricting how aggressively the GPU requests voltage under load. That could potentially reduce the likelihood of overheating connectors or damaging hardware.
The trade-off, however, is clear: if voltage is being restricted too aggressively, GPU boost behavior and overclocking headroom can take a hit. For enthusiasts who buy premium GPUs specifically for top-tier performance—and in some cases for overclocking—this kind of driver change can feel like a performance “nerf,” even if it’s meant to protect the hardware.
For now, users on RTX 5090 and other potentially affected RTX 40/50 series cards may want to pay close attention to voltage, clock speeds, and performance benchmarks after updating to driver 595.71. If performance is lower than expected or overclocks no longer scale the way they used to, rolling back to a previous stable driver may be a temporary workaround until Nvidia clarifies whether this is an intentional safety measure or an unintended bug that will be corrected in a future update.
