Apple plans to power a revamped version of Siri using Google's Gemini AI model and Nvidia hardware [1, 2].
This partnership signals a significant shift in Apple's AI strategy by relying on external cloud infrastructure and third-party models to compete in the generative AI race. By integrating these technologies, Apple aims to improve the conversational and search capabilities of its virtual assistant [5, 1].
According to reports, the updated Siri will run on Nvidia Blackwell B200 GPUs [4, 6]. These chips will be hosted on Google Cloud's data-center infrastructure [4, 6]. The company intends to use a custom version of the Gemini AI model specifically tailored for Apple's ecosystem [2].
Security remains a primary focus of the integration. Apple intends to employ Nvidia's confidential-computing features to encrypt user data during processing [5, 1]. This approach is designed to maintain user privacy while utilizing the high-compute power of the B200 GPUs [4].
There are conflicting reports regarding the release timeline. Some sources said the revamped Siri could debut as soon as next week [4]. Other reports said the overhaul may arrive later with the release of iOS 27 [6].
The move represents a complex technical alliance between three of the world's largest technology firms—Apple, Google, and Nvidia—to bridge the gap in AI performance.
“Apple plans to power a revamped version of Siri using Google's Gemini AI model and Nvidia hardware.”
This arrangement demonstrates Apple's pragmatic approach to the AI era, choosing to integrate industry-leading hardware and software rather than relying solely on proprietary internal development. By utilizing Google Cloud and Nvidia's B200 chips, Apple can deploy a more powerful Siri faster, while the use of confidential computing attempts to preserve its brand promise of privacy.
