NVIDIA imposes its ACPI standard in Linux 7.2 for its own interests

Published on June 08, 2026 | Translated from Spanish

The next version of the Linux kernel 7.2 will integrate support for the ACPI CPPC v4 standard, created by NVIDIA. This technology allows the operating system to communicate to the hardware when performance is normal, seeking to avoid overheating or excessive consumption. However, behind this seemingly altruistic collaboration lies a strategy of colonizing free software with proprietary standards.

NVIDIA logo-shaped motherboard chip glowing intensely while a Linux kernel gear mechanism is slowly wrapped by metallic ACPI CPPC v4 chains, system overheating warning indicators flash red around a processor die, a robotic hand tightening proprietary lock clamps onto open-source circuit traces, cooling fan blades struggle against viscous golden liquid dripping from the chip, cinematic engineering visualization, dark server room background, dramatic orange-blue lighting contrast, photorealistic technical illustration, macro lens focus on chip surface details, smoke wisps rising from overheated VRM components

Open source, but closed control of thresholds 🔒

Although the source code of ACPI CPPC v4 is available, the fine parameters that define normal performance are exclusively set by NVIDIA. The user will not be able to modify these thresholds without risking warranties or stability. This especially benefits NVIDIA's data centers, where their chips optimize their efficiency, while the competition could see inferior performance by not being able to adapt to this partially closed standard. The kernel becomes larger and more complex to solve a problem that the hardware should manage on its own.

The hardware takes a coffee break while the kernel does its work ☕

So now it turns out that your processor, instead of self-regulating as before, needs the kernel to tell it whether it's working normally or overheating. It's like your refrigerator asking for permission to cool down. And the best part: NVIDIA defines what is normal, so if your AMD graphics card heats up a bit, the kernel will tell it to slow down while the RTX cards keep running at full speed. All for the common good, of course.