Google will implement requirements for Android applications to have optimized code to consume less RAM, establishing specific limits for this resource within the operating system. The company's justification for this new measure is the current crisis in the supply of memory chips.
Applications that do not adhere to the new guidelines may face technical sanctions—such as app crashes or closures—as well as distribution-related penalties, resulting in reduced visibility on the Play Store. These policies were formally announced on Wednesday, August 26th, and will take effect in February 2027.
The company will define distinct limits according to the type of memory usage. Dynamic usage covers the storage of application data, whether active or compressed, while memory usage dedicated to bitmaps is linked to displayed graphic elements.
The limit values are adjusted according to the physical RAM capacity of the device and the state in which the application is. For dynamic usage, a minimum threshold is set for devices with 4 GB of RAM, where the app must not exceed 2 GB in the foreground or 1 GB in the background. In contrast, the higher capacity category, with 16 GB of RAM, will have limits of 4.25 GB in the foreground and 2 GB in the background.
For games, the limits are higher: 2.25 GB on 4 GB devices and 5 GB on 16 GB models. Regarding bitmap memory usage, the limits only apply when applications are in the background. This makes sense because, being a visual use, removing the app from the screen should decrease the allocated space; in this scenario, consumption must not exceed 200 MB in the background or 400 MB when cached.
Additionally, regarding code optimization, Google will require a 25% reduction in DEX code, which can be achieved by eliminating unused code, shortening instructions, or reducing class names.
In recent years, large investments in building data centers have led certain corporations to acquire a significant share of memory chip production. Consequently, major suppliers have concentrated all manufacturing on components used in artificial intelligence training and inference. This has left little supply for other sectors of the electronics industry, and the little available has become much more expensive. The ramifications of this scenario are diverse, including underperforming devices, sales decline, increased product costs, production line shutdowns, extinction of categories, and record profits for chip manufacturers.
Google explains that this situation transcends mere contexts, pointing out that: 'The mobile device industry is undergoing significant hardware supply restrictions, which are altering memory availability, which, over time, can negatively impact user experience.'
As noted by The Verge, Google itself faced this reality in its smartphone line, since the Pixel 11 Pro and Pixel 11 Pro XL, launched in August 2026, have less RAM than their previous generations. Thus, the company cannot exempt itself from responsibility, given that it must invest approximately US$ 200 billion in AI infrastructure.
