Apple will run its frontier language model FM Cloud Pro on Nvidia GPUs via Google Cloud, departing from its private M-series infrastructure to handle agentic AI workloads.

Apple software chief Craig Federighi confirmed that FM Cloud Pro, the company's new frontier language model built to be on par with Google's Gemini frontier models, will run on Nvidia GPUs hosted in Google Cloud for agentic workflows (AI systems that autonomously execute multi-step tasks) and complex reasoning.

Apple's private cloud infrastructure running on M-series Mac chips could not sustain the scaling demands of the new agentic experience after two years of operation, prompting the shift to outsourced compute despite Apple's historical emphasis on privacy-first, vertically integrated infrastructure.

Nvidia's Data Center revenue hit $75.25 billion, up 92% year over year, and the company guided next-quarter revenue to $91.0 billion plus or minus 2%; Google Cloud revenue grew 63% to $20.03 billion in Q1, with backlog nearly doubling quarter over quarter to over $460 billion.