Apple Intelligence, a new personalized AI experience, requires significant computational power to run large language models efficiently. The iPhone 15 Pro and iPhone 15 Pro Max models, as well as Macs and iPads with M1 or later chips, have the necessary hardware capabilities, including powerful Neural Engines and ample RAM, to support this feature. Older devices may not have the required performance to run these models at useful speeds, which is why they are not compatible with Apple Intelligence.
John Giannandrea explained that the computational demands of running the large-language models used by Apple Intelligence are significant due to the inference process being computationally expensive. Factors contributing to these demands include the device's bandwidth, the size of the Apple Neural Engine, and the device's overall processing power. While it would be theoretically possible to run these models on older devices, the slow speed would render them practically unusable.
Apple Intelligence, introduced with iOS 18, iPadOS 18, and macOS Sequoia, is a personal intelligence system that combines the power of generative models with personal context to deliver highly useful and relevant intelligence. It enhances the user experience across iPhone, iPad, and Mac by understanding and creating language, taking action across apps, drawing from personal context, and simplifying everyday tasks. Built with privacy from the ground up, Apple Intelligence is deeply integrated into these operating systems, harnessing the power of Apple silicon to provide an advanced AI experience.