According to recent reports, Apple plans to power its new Siri on NVIDIA Blackwell chips through Google Cloud starting in September, just days after NVIDIA launched its most powerful AI model yet.
We break down the potential alliance, NVIDIA’s freshly unveiled Nemotron 3 Ultra, and what this could mean for the broader AI race.
Why Apple Could Lock In a Major NVIDIA Deal
Apple plans to launch a new generation of Siri in September 2026, and several reports from The Information confirm the assistant will rely on NVIDIA chips behind the scenes for cloud-based AI processing tasks.
The setup is three-way. Apple will run as much processing as possible on-device, but heavier queries will flow to Google Cloud through a licensed version of Gemini. That cloud infrastructure runs on NVIDIA Blackwell B200 data center chips.
According to reports, Apple had recently approved NVIDIA’s confidential computing technology. The feature encrypts data and AI models while they are processed on the chips, allowing Apple to keep its privacy standards while using external cloud servers for advanced functions.
This is significant for both companies. Apple gets access to far more compute than its Private Cloud Compute alone could provide. NVIDIA effectively becomes critical infrastructure for one of the largest consumer AI launches in years.
Follow us on X to get the latest news as it happens
The arrangement also strengthens NVIDIA’s position against rivals. Apple, Google, and NVIDIA together form one of the most powerful AI stacks in consumer technology, with the Blackwell B200 designed for large-scale model training and fast inference.
Investors will watch closely. WWDC 2026 begins on June 8, and Apple is expected to outline its full AI strategy. If the integration delivers, NVIDIA could see its enterprise AI footprint expand sharply across the most demanding consumer applications.
Share prices for both companies saw slight increases amid the reports. NVDA was trading at $216.18 after rising 0.71% in the last 24 hours. Meanwhile, APPL traded at $310.04, up 0.2% over the same period, according to TradingView data.
NVIDIA Unveils Nemotron 3 Ultra: Its Most Powerful AI Model to Date
Nemotron 3 Ultra is NVIDIA’s new open-source AI model with roughly 500 to 550 billion parameters. CEO Jensen Huang presented it at Computex 2026 in Taipei on June 1, designed for advanced reasoning and complex agentic workflows.
An agentic workflow is an AI system that plans, executes, and iterates on multi-step tasks with minimal human oversight. Nemotron 3 Ultra sits at the top of a three-tier family that also includes the Nano and Super variants.
“Nemotron 3 Ultra is built for that new workload. It’s a frontier smart model that delivers up to 5x faster inference and lowers the cost of complex agentic tasks by up to 30%. This enables agents to finish the same job in less time or complete more jobs in the same time,” NVIDIA said.
Adoption is already strong. The Nemotron 3 family recorded more than 50 million downloads in the year leading up to April 2026, signaling that open model strategy is working among developers and enterprise customers worldwide.
For enterprise users, the 5x throughput improvement matters because it sharply lowers cost-per-inference. That dynamic positions NVIDIA not just as a chipmaker, but as a full-stack AI platform company capable of competing directly with closed model providers.
The timing also matters. With Apple set to lean on NVIDIA hardware for Siri and Nemotron 3 Ultra reinforcing its software credibility, NVIDIA is sealing both ends of the AI stack precisely when the next consumer cycle begins.