Apple’s use of Nvidia chips for Siri shows a new reliance on external AI infrastructure, useful context for a colleague following privacy and cloud shifts.
Apple taps Nvidia for AI-powered Siri Story flow and key facts
Apple is planning to use Nvidia’s Blackwell B200 data center GPUs to power certain queries for its upcoming Gemini-integrated Siri, according to a report from The Information. These computations will occur in Google Cloud, marking a notable shift from Apple’s traditional approach of tightly controlling its hardware stack. The move highlights Apple’s need for external AI infrastructure to support advanced on-device and cloud-based intelligence.
To address privacy concerns, Apple has approved the use of Nvidia’s confidential compute technology, which encrypts data while it’s being processed on the GPU. This hardware-based security feature ensures that sensitive user information remains protected even in shared cloud environments. The Blackwell B200 chips, part of Nvidia’s latest architecture, are designed for large-scale AI inference and training, offering significant improvements over previous generations.
The integration raises questions about how this cloud-dependent processing will align with Apple’s existing Private Cloud Compute system. While Apple maintains its commitment to on-device processing for many tasks, this partnership signals a strategic reliance on external cloud and chip partners to deliver next-generation AI features at scale.
Facts
- Apple will use Nvidia’s Blackwell B200 data center chips via Google Cloud for some Siri queries.
- Nvidia’s confidential compute feature will encrypt data during AI processing to preserve privacy.
- The Blackwell architecture succeeds Hopper and supports trillion-parameter AI models.
- Apple’s use of external cloud infrastructure contrasts with its usual strategy of controlling core hardware components.
Canto visual news explainer. AI tools may assist production. Editorial policy
