Apple Is Using Google and Nvidia to Fix Siri

Apple building its most important features in-house is practically a brand identity at this point. Which is exactly why this story matters so much. Reports confirm Apple is preparing the biggest Siri overhaul in years and to get it done, the company is turning to Google’s Gemini AI models, Google Cloud infrastructure, and Nvidia Blackwell B200 chips. Three external partners for a product Apple has always treated as its own.

The architecture is a deliberate hybrid. Apple handles on-device processing where its own silicon can manage it. Google’s Gemini models provide the conversational intelligence and language processing foundation. Google Cloud supplies the server infrastructure. And Nvidia’s Blackwell B200 AI chips handle the heaviest inference workloads inside those data centers selected apparently because Apple’s own Private Cloud Compute struggled with large-scale AI processing demands during internal testing.

The Google partnership isn’t entirely new Apple and Google confirmed a multi-year agreement centered on Gemini-powered Apple Foundation Models earlier this year. What’s new is the Nvidia layer. Apple reportedly chose Nvidia’s hardware specifically for its confidential computing capabilities technology that encrypts data while it’s actively being processed inside cloud servers. That’s Apple trying to maintain its privacy-first reputation while simultaneously outsourcing the computer it couldn’t build fast enough itself.

What does the new Siri actually do? Personal context awareness remembering what’s happening across your apps and life. On-screen understanding. Deeper app integration for multi-step task automation. More natural conversations with proper follow-up question handling. Chatbot-style interactions. Features Apple announced at WWDC 2024 that got delayed, then delayed again, now arriving properly.

The September 2026 rollout alongside new iPhones is the current target though exact timing on specific features remains uncertain.

The competitive context is straightforward. Google Gemini and Samsung Galaxy AI have been delivering these capabilities for over a year. Apple is catching up by using its competitors’ infrastructure to get there faster rather than waiting another year for internal solutions.

Source