Apple is set to unveil a revamped Siri in September, built on AI hardware and models from Nvidia and Google. The tech giant will lean heavily on Nvidia's Blackwell chips, while the voice assistant taps into Google's Gemini model hosted on Google Cloud. This marks a rare move for Apple to outsource core AI processing, promising richer interactions for millions of users.

Apple historically keeps AI development internal, but the partnership promises faster rollout of advanced features like contextual understanding and on‑device privacy safeguards. Queries routed to Google Cloud will run on a licensed Gemini instance, while Nvidia's Blackwell B200 silicon adds confidential compute that encrypts data during processing. The hybrid approach aims to balance performance with Apple’s privacy stance, while maintaining compliance with global data regulations.

Consumers can expect a Siri that responds faster and handles more complex requests, but some interactions will leave Apple’s ecosystem for third‑party servers. By leveraging Nvidia’s hardware and Google's model, Apple sidesteps the need to build its own large‑scale AI chips, accelerating the timeline for a competitive voice assistant, and could set a new standard for privacy‑first AI services across mobile platforms.