News
IndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value ChainIndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value Chain
EveryChip
Neural Engine

Neural Engine

Apple's on-device AI accelerator for machine learning tasks

Neural Engine is a series of artificial intelligence accelerators built into Apple system-on-chips. Introduced in 2017 with the A11 Bionic, it powers features like Face ID and Siri while keeping user data private through local processing.

Share on WhatsApp

  • First introduced in 2017 with the A11 Bionic chip
  • M4 model performs 38 trillion operations per second
  • Enables on-device processing for Face ID and Siri
  • Integrated into all A-series and M-series processors since 2017
  • Supports Apple Intelligence features like Image Playground

What it is

Neural Engine is a dedicated hardware component designed by Apple to accelerate machine learning workloads. It first appeared in 2017 inside the A11 Bionic chip. This accelerator handles complex mathematical operations required for artificial intelligence tasks directly on the device. By processing data locally rather than sending it to remote servers, it ensures user privacy and reduces latency. Every A-series and M-series processor released since its debut includes this technology. It supports real-time applications such as facial recognition, voice assistance, and augmented reality experiences.

How it came about

Apple introduced the Neural Engine alongside the A11 Bionic system-on-chip in 2017. The initial deployment targeted the iPhone 8, iPhone 8 Plus, and iPhone X models. In 2020, the technology expanded to personal computers when Apple launched the M1 processor for Mac devices. Since then, the company has integrated the accelerator into all subsequent A-series and M-series chips. This consistent inclusion marks a strategic shift toward embedding AI capabilities directly into consumer hardware across phones, tablets, and laptops.

Inside the chip

The architecture focuses on high energy efficiency to perform intensive AI calculations without draining battery life quickly. It works closely with Apple Core ML framework, allowing software developers to deploy machine learning models that run on-device. This setup supports functions like object recognition, natural language processing, and gesture detection. Recent iterations show significant performance gains. The version in the M4 chip can execute 38 trillion operations per second, nearly doubling the 18 trillion operations per second capability found in the preceding M3 model.

How it was made

Apple designs the Neural Engine as part of its broader silicon strategy. While specific manufacturing details for the accelerator alone are not isolated in public records, it is fabricated as part of the larger system-on-chip. These chips are produced using advanced semiconductor processes. The integration allows the AI unit to share resources efficiently with central and graphics processing units within the same package, optimizing both space and power consumption for mobile and desktop environments.

Where it went

This hardware powers several key user-facing features. Face ID uses it for secure facial recognition, while Siri relies on it for voice command processing. Computational photography tools like Smart HDR and Night Mode leverage the engine to enhance images in real time by processing sensor data. Augmented reality applications also benefit from its speed. In 2024, Apple Intelligence services expanded its role, enabling features such as Image Playground for generating pictures and Writing Tools for text correction on iPhones, iPads, and Macs.

Versions and family

The Neural Engine has evolved through multiple generations of Apple silicon. It began with the A11 Bionic and continued through every A-series chip for mobile devices. The M1 brought the technology to Mac computers, followed by the M2, M3, and M4 families. Each generation typically offers improved throughput and efficiency. The M4 variant represents a notable leap in raw performance compared to the M3, reflecting ongoing refinements in the underlying design to handle increasingly complex AI workloads.

Why it matters

By moving AI processing to the device, Apple has set a standard for privacy-centric machine learning. Users can enjoy intelligent features without compromising sensitive data, as information does not need to leave the handset or computer. This approach has influenced how consumers view security in smart devices. The widespread adoption across billions of Apple products has also encouraged developers to optimize applications for on-device inference, creating a robust ecosystem for local AI execution.

Questions people ask about Neural Engine

What is the Neural Engine used for?

It accelerates machine learning tasks such as Face ID, Siri voice recognition, and computational photography. It also supports augmented reality and new Apple Intelligence features like text proofreading and image generation.

Does the Neural Engine protect user privacy?

Yes. By processing data locally on the device instead of sending it to cloud servers, it keeps sensitive information like facial scans and voice commands secure and private from external access.

Which Apple chips include a Neural Engine?

Every A-series and M-series processor released since 2017 includes this accelerator. It started with the A11 Bionic and is now present in the latest M4 chips for Macs and A-series chips for iPhones.

Good reading and sources

Where this page comes from

The facts on this page come from the Wikipedia article "Neural Engine", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryChip rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.

It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

Photo credits on this page

The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

The weekly digest

Every Monday: the chip of the day, new pages on EveryChip, and the conferences and anniversaries of the week ahead.