New Siri to run advanced AI on Google servers
Apple is partnering with Google to power its next-generation Siri using Gemini models and cloud servers equipped with Nvidia’s Blackwell B200 chips. Internal testing showed Apple’s Private Cloud Compute was too slow for advanced AI workloads, prompting the shift to Google’s infrastructure. To maintain privacy, Apple will leverage Nvidia’s confidential computing technology, encrypting user data during processing. Basic Siri requests will remain on-device, while complex queries will rely on this hybrid cloud approach, set to be unveiled at WWDC.
Read the full story on Macworld →