Image: Denny Müller/Unsplash
Google rebuilt Android’s kernel using AutoFDO and real app data, boosting performance, speeding system calls, and improving app launches.
Android could soon feel noticeably faster after Google rebuilt part of the operating system’s core using real-world usage data.
The change relies on Automatic Feedback-Directed Optimization (AutoFDO), a technique that analyzes how software actually runs on devices and reorganizes code to prioritize the tasks users trigger most often. Instead of relying on theoretical compiler assumptions, Android’s core can now be optimized around real app behavior.
To power the update, Google engineers gathered performance data from tests using the 100 most-used Android apps on Pixel devices. That data helped identify the most frequently executed parts of Android’s core software, allowing engineers to rebuild sections of the system so those “hot” code paths run faster and more efficiently.
Google didn’t add a new feature to Android, and users aren’t expected to do anything new. But phones will be faster, and users will notice the difference in performance. What happened then?
Instead of bringing a new feature, especially one that aligns with the AI spree, it simply modified Android code a bit.
The new approach targets Android’s core: the kernel. Because the kernel accounts for around 40% of CPU time, according to Android Developers, optimizing its task handling delivers a noticeable, device-wide speed increase.
To gather the necessary optimization data, Google stopped relying on inaccurate static data and focused on the Android compiler. The team ran usage tests using the 100 most-used apps, identifying the core software segments used most often. They rebuilt the kernel using this data. Now, these frequently used tasks, known as “hot-code,” run more efficiently, enabling faster app launches, smoother multitasking, and better battery life.
The initial work occurred in a lab. For continuous, real-time optimization, AutoFDO takes over. AutoFDO collects data and dynamically reorganizes the code on the user’s device, prioritizing CPU resources based on frequent user actions
By optimizing the Android kernel code for speed prioritization, Google was able to improve the following:
According to TechRader, this update is being tested on Android versions 15, 16, and 17 beta. It suggests that only devices running these versions will receive the optimization when their manufacturers ship kernel updates that include the changes.
As the optimization rolls out, older phones may become eligible, since they seem to need it more. The exact date the rollout will start isn’t known yet, but based on how Google ships updates, Pixel phones will likely get it first.
For more on where the platform is headed next, check out our breakdown of the biggest Android trends shaping smartphones in 2026.
Joseph is a Technical Writer with about 3 years of experience in the industry, also advancing a career in cyber threat intelligence. He is passionate about the responsible use of technology, a passion that led him into cybersecurity. As an undergrad, he leads a novel community of technology enthusiasts at his school, NOUN, where he guides and shares resources for beginners in tech. His writing experience includes writing on a diverse range of topics, from consumer tech to startups and tutorials. Additionally, he periodically shares case studies and research reports on cybersecurity on his social media pages.
