New iPhone 16 leak reveals the special speed boost underpinning the A18 chip

After months of leaks, Apple will announce the iPhone 16 family today. A major draw of the new phones will be their AI capabilities. Since many of the features will run on-device, the phones will need to have a lot of oomph, and for that, they will require a fast chipset. We already know a fair amount about the A18 chip that will power the new phones and a report has now revealed that it will be based on Arm’s newest V9 architecture.

Arm's chip designs specify the instructions that influence how a chip works. Apple's fastest chip for iPhones, the A17 Pro, which powers the iPhone 15 Pro and Pro Max, is based on the previous-gen V8 design.

Apple's M4 chip that fuels the latest iPad Pro is also based on the Arm V9 and is a considerable leap over older chips. The new architecture technically allows for up to a 30 percent performance boost. It also comes with enhanced AI and machine learning features.

Apple AI offerings fall under the Apple Intelligence umbrella and will be available as part of iOS 18. The only two current iPhones that will be able to run the features are the iPhone 15 Pro and Pro Max, which happen to be one of the best phones of 2024 and are fast enough to run the models behind these AI functionalities.

The AI capabilities are similar to what we have already seen from other companies such as Google and Samsung and will let you summarize text, create new images based on prompts, and get more out of Siri.

The new A18 chip is rumored to be around 10 percent faster than the A17 Pro, which might not sound like a lot, but considering the A17 Pro is already pretty snappy, it's a decent enough bump.

The chip's Neural Engine, which is a group of specialized cores for AI tasks, is said to have more cores than the A17 Pro's Neural Engine. The chip is also rumored to have a larger die area for edge AI computing, which means it will be able to fit in more components for increased performance. This will help with edge AI or on-device AI processing.