Apple reportedly turning to Nvidia chips for Gemini-powered Siri

Apple will reportedly lean on Nvidia to fuel the next generation of Siri.

The Information reported that Apple will "tap into Google’s fleet of Nvidia’s Blackwell B200 data center chips" to power requests from the new AI-enhanced Siri, which will allegedly go online with the launch of iOS 27 later this year. It's already been established that the new Siri will be based on Google's Gemini AI model, so Apple utilizing a Google fleet of Nvidia data center chips makes sense.

In layman's terms, the main thing to know about this is that Nvidia's Blackwell architecture allows for large-scale AI usage with high degrees of memory bandwidth and other important things you need to make a popular chatbot work. These chips also have the ability to encrypt data that comes through them, which would help maintain Apple's high privacy standards as it attempts to catch up to the rest of the tech industry with AI features on its devices.

The new Siri hasn't officially been shown off yet, but it's widely expected that it will exist in the form of a chatbot with its own dedicated app once iOS 27 launches. It will be able to remember personal context for the user as well as read anything that's on screen, giving it capabilities far beyond what Siri has been able to do previously.

We should find out come Monday, when Apple hosts its annual developer conference, WWDC 2026.