Open-source boot utility lifts AI processing limits on Nvidia GeForce cards

A community-developed open-source project named nv-uefi-unlocker allows owners of Nvidia GeForce RTX 20, 30, and 40-series graphics cards to bypass built-in Tensor Core restrictions on Windows and Linux. The utility boots from a FAT32 USB drive prior to operating system initialization, lifting throttles on tensor processing units without flashing video BIOS or modifying drivers. In benchmarks highlighted by Tom's Hardware, a laptop RTX 4060 achieved 65 percent higher FP16 math performance, while Wccftech notes the utility promises up to 75 percent higher AI compute performance across supported hardware. [1][2]

The utility works by executing a payload that unlocks speed-selector registers in Nvidia's signed Booter firmware, followed by read-back verification before passing control to the operating system loader. Because the modification does not persist through a complete system power-off, users must execute the boot process every time they turn on their computers. Systems must run in UEFI mode with Secure Boot disabled or configured to trust a self-signed EFI file, a requirement that cannot bypass certain anti-cheat software in games. [1][2]

Hardware compatibility varies across generations. While the tool supports desktop and laptop models spanning the RTX 30 and 40 series, specific cards such as the RTX 2080, RTX 2080 Ti, and RTX 2080 Super encounter startup failures. RTX 20-series users must also switch their drivers into GSP firmware mode beforehand. Pre-Turing graphics cards and upcoming RTX 50-series hardware remain unsupported, with the development team explaining that newer Blackwell hardware incorporates substantially tighter security measures. [1]