Apple Shifts Siri AI to Nvidia Blackwell Chips on Google Cloud

• Apple adopts Nvidia B200 processors on Google Cloud to power new Siri AI features.
• Move boosts performance and privacy ahead of WWDC 2026 and iOS 27 launch.

On June 4, 2026, The Information reported that Apple is replacing its in-house infrastructure for Siri’s new AI features with Nvidia Blackwell B200 processors located in Google Cloud data centers. Apple made this decision after identifying performance constraints with its Private Cloud Compute system, which previously used Mac-series silicon for large language models. By switching to external Nvidia hardware, Apple aims to deliver faster and more competitive generative AI capabilities within Siri.

Apple is using Nvidia’s confidential compute technology to protect user data during cloud-based processing on Google servers. This approach uses hardware-level encryption to address privacy concerns in the new system. Sources noted that Apple is likely to keep the Private Cloud Compute branding, despite the backend now running on third-party infrastructure.

According to MacRumors, the shift highlights Nvidia’s ongoing dominance in AI cloud infrastructure and Apple’s reliance on top external hardware to match competitors like ChatGPT and Claude.

Yahoo Tech described the revamped Siri as including context-aware responses, the ability to analyze on-screen content, and a dedicated application. Apple plans to preview these features at WWDC 2026 on June 8. The public rollout of the updated Siri is set for September, alongside iOS 27.