3.The story
- 2008 Johny Srouji joins to lead the A4, the first Apple-designed system-on-a-chip 16.
- 2017 A11 Bionic in iPhone X brings the first Neural Engine, for Face ID 3.
- 2020 M1 moves the Mac to Apple silicon 1.
- 2024 M4 launches; Apple Intelligence arrives on iPhone, iPad and Mac 13 19.
- 2025 M5 puts an AI unit in every graphics core; Houston-built AI servers start shipping 4 20.
- 2026 Google Gemini deal, M5 Ultra, and the first 2 nm chips, M6 and A20 Pro 21 6 2.
The first Neural Engine managed 600 billion operations per second 3; M1's did 11 trillion 1 and M4's 38 trillion, which Apple called "60x faster than the first Neural Engine" 13. From 2025 Apple stopped quoting a single figure and instead compares each chip with the one before: M5 has "over 4x the peak GPU compute performance for AI" of M4 4.
Apple announced Private Cloud Compute in June 2024 as "custom-built server hardware that brings the power and security of Apple silicon to the data center" 5. In January 2026 Apple and Google said "the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology", while Apple Intelligence "will continue to run on Apple devices and Private Cloud Compute" 21.