Key facts
- Apple reportedly plans to use Nvidia's Blackwell B200 chips for its next-generation Siri.
- The new Siri will integrate Google's Gemini AI model.
- Nvidia's B200 chips offer high memory bandwidth and data encryption.
- The enhanced Siri is expected to function as a chatbot with advanced features.
- This represents a potential compromise on Apple's initial plan to use only its own silicon for AI tasks.
Apple is reportedly planning to utilize Nvidia's Blackwell B200 data center chips to power its next-generation Siri, which will be integrated with Google's Gemini AI model. This strategic move aims to equip Siri with advanced AI capabilities, with the enhanced version expected to be unveiled alongside iOS 27 later this year. The Blackwell architecture is recognized for its substantial memory bandwidth, essential for demanding AI applications such as chatbots. Furthermore, the chips' integrated data encryption features are intended to help Apple uphold its stringent privacy standards. The forthcoming Siri is projected to operate as a dedicated chatbot, offering enhanced functionalities including context memory and on-screen content reading, significantly expanding its current capabilities. The report originates from The Information, and Apple has not officially announced these plans. The article suggests this represents a compromise for Apple, which had previously aimed to use its own silicon for such AI tasks, but is now relying on Google's servers and Nvidia's hardware for optimal inference and to maintain some privacy credentials.