The iPhone is constantly evolving, and even though the iPhone 17 series hasn't been released yet, initial information about the next generation is already available. The A20 chip, which is expected to be used in the iPhone 18 models, will remain on 3nm technology but will receive a significant upgrade. Apple is relying on a new packaging technology that could, in particular, improve the performance of Apple Intelligence. This potentially means a faster and more efficient iPhone with better artificial intelligence.
Apple has made consistent progress in developing its own chips in recent years. This not only involves increased computing power, but also energy efficiency and optimization for new features. What's particularly striking about the A20 chip is that Apple has decided against switching to 2nm technology and instead relied on improved 3nm manufacturing. The key factor here is the integration of components, which enables closer collaboration between the processor, memory, and neural engine.
A20 chip stays at 3nm, but gets a crucial upgrade
The A20 chip, planned for the iPhone 18 series, will be manufactured using TSMC's third-generation 3nm process. This manufacturing technology, known as N3P, is also expected to be used in the A19 and A19 Pro chips in the iPhone 17 models. Those hoping for a major performance leap may be disappointed, as the increase in raw power is likely to remain manageable due to the same manufacturing technology. However, the A20 chip is expected to feature an important innovation. Apple is using TSMC's Chip-on-Wafer-on-Substrate (CoWoS) technology for the first time. This ensures tighter integration of the processor, RAM, and Neural Engine, which could particularly benefit the iPhone's AI capabilities.
Better performance for Apple Intelligence
Apple Intelligence is Apple's proprietary artificial intelligence, which is becoming increasingly important in iOS, macOS, and other Apple products. With the new packaging technology of the A20 chip, Apple could significantly improve its AI capabilities. CoWoS technology enables more efficient collaboration between the processor and the Neural Engine. This could enable AI tasks such as speech processing, image recognition, and machine learning to be performed faster and more energy-efficiently. This could bring tangible benefits in the following areas:
- Improved Siri performance with more natural speech recognition and faster response times
- More efficient photo and video processing through optimized image analysis
- Faster and more accurate text recognition in apps like Notes and Safari
- Better battery life through optimized AI processes
These improvements are particularly relevant as Apple increasingly relies on artificial intelligence in the coming years. The A20 chip could be a crucial step toward making the iPhones of the future even smarter and more efficient.
When will 2nm technology arrive?
Apple will not switch to 2nm manufacturing with the A20 chip. According to analyst Jeff Pu, this switch will only take place with the A21 chip, which will be used in a new iPhone in 2027 at the earliest. Until then, Apple will stick with 3nm technology, but will focus on optimizations through new packaging techniques and improved integration of chip components (via GF Securities).
iPhone 18: Apple focuses on optimization instead of new manufacturing technology
The A20 chip in the iPhone 18 series won't bring a revolution in computing power, but it will represent a significant upgrade for Apple Intelligence. Thanks to CoWoS technology, Apple could significantly improve its AI capabilities and make the iPhone even more powerful. Those waiting for 2nm production will have to wait until 2027. So, for now, Apple is focusing on optimizations rather than a complete technological leap. (Photo by Yurchanka / Bigstockphoto)
- Apple optimizes foldable display for greater efficiency
- iPhone 17: Apple installs its own Wi-Fi 7 chip for the first time
- iPhone 17 Air with MagSafe? Leaker reveals new details
- iPhone 18 Pro: Apple relies on its own development for the modem
