Apple is developing custom silicon for future products including AR glasses, wearables, and cloud AI infrastructure, with plans for successive generations like M6 and M7 chips and custom components such as 5G modems.

Apple's strategic shift to designing its own custom silicon, initiated over a decade ago and intensified with the launch of M-series chips in 2020, has significantly transformed the global semiconductor market and the AI landscape.

This revolution in silicon design is driving a broader industry shift toward vertical integration, specialized hardware, and on-device AI, which could reshape global supply chains, geopolitical strategies, and the future of AI development.

Apple’s silicon strategy signifies a move towards more resilient, efficient, and privacy-centric AI development, with profound implications for supply chains, geopolitics, and AI-powered devices and services.

Apple's A-series chips, starting with Neural Engines in 2017, introduced AI hardware that powers features like Face ID and computational photography, with the latest A18 and A18 Pro chips featuring a 16-core NPU capable of 35 TOPS.

The latest chips in the A-series, such as the A18 and A18 Pro, include advanced AI accelerators like a 16-core Neural Processing Unit, significantly boosting AI capabilities on Apple devices.

Under Project 'Baltra,' Apple is creating dedicated AI server chips in partnership with Broadcom to enhance control over its AI infrastructure and support large models, reducing reliance on third-party hardware.

Apple's custom silicon involves designing highly integrated System-on-a-Chip (SoC) components that combine CPU, GPU, Neural Engine, and memory on a single chip, primarily manufactured by TSMC using advanced 3nm and 5nm processes.

The M-series chips, starting with M1 in 2020 and including the latest M5 in 2025 built on 3nm technology, have significantly boosted performance and AI capabilities, featuring over four times GPU performance and a 16-core Neural Engine.

This silicon revolution has benefited TSMC through increased demand for cutting-edge process nodes like 3nm and 2nm, while challenging traditional chipmakers such as Intel, Qualcomm, AMD, and Nvidia to accelerate their AI hardware development.

Apple's focus on on-device AI and privacy-first computing promotes edge computing, reducing reliance on cloud services, and could redefine user expectations for privacy, responsiveness, and offline AI capabilities.