Apple is exploring acquisitions of AI‑chip specialists to reduce its reliance on Nvidia, people close to the matter say. Today, the company compensates for limitations on the heaviest AI workloads by leaning on Nvidia hardware hosted by cloud partners — a setup Apple reportedly wants to move away from over the long term.
Why Apple is targeting AI‑chip companies
Apple already builds chips for its data centers, but those processors — originally engineered for mobile and Mac workloads — have shown limitations with some large AI models. During work on the new Siri, teams reportedly had to offload parts of the Gemini model’s compute to Nvidia gear in the cloud. Buying outside expertise in AI accelerators would enable Apple to develop purpose‑built on‑premises hardware for large‑model inference and training, reduce cloud dependency, improve latency and energy efficiency, and achieve tighter integration with its software and silicon roadmap.
