
The chip that ate the world
From Design to Dominance: ARM’s Bold Leap into AI Chip Manufacturing
(This article was generated with AI and it’s based on a AI-generated transcription of a real talk on stage. While we strive for accuracy, we encourage readers to verify important information.)
ARM, a foundational technology company, has strategically shifted from designing computer chips to manufacturing its own. Mr. Will Abbey, Executive Vice President and Chief Commercial Officer at Arm Ltd., explained this move is driven by evolving AI infrastructure demands, broadening ARM’s market impact.
Mr. Abbey highlighted ARM’s core focus on performance and power efficiency, crucial for future AI agentic workloads. This expansion enables companies to access ARM technology not just as intellectual property (IP) or compute subsystems, but now also as complete chips, offering greater flexibility and tailored solutions.
Addressing potential competition with partners like Nvidia, Microsoft, and Google, Mr. Abbey clarified ARM’s new approach provides “optionality.” The industry needs more solutions for AI infrastructure complexities. ARM will continue collaborating, offering IP and subsystems, while its own chips serve as an additional pathway.
Regarding AI usage measurement, Mr. Abbey advocated focusing on measurable business outcomes over token consumption. Tokens are inputs, and their cost is rising. Effective measurement should consider factors like time saved, new features, and bug rates, aligning AI deployment with tangible business objectives.
Power efficiency has been central to ARM since its inception 35 years ago, initially for mobile phones. This core sensibility is now critical in the AI era. The industry recognizes that simply increasing performance or data centers is unsustainable. Efficient system architecture is paramount, complemented by renewable and nuclear power.
ARM’s new AGI silicon, a data center chip for future AI agentic workloads, has seen remarkable demand. Forecasted revenue for fiscal years 2027-2028 has doubled to $2 billion. However, securing sufficient supply of components like memory, PCBs, and substrates remains a challenge. ARM is actively engaging its ecosystem to meet this astronomic demand.
The discussion also touched upon the geopolitical significance of AI chips. Mr. Abbey emphasized that AI is a powerful technology requiring sensible deployment with security at its core. He stressed collaboration between industry and government to ensure policies protect sovereign data and enable safe innovation.
As a fundamental IP company, ARM prioritizes intellectual property protection. While ARM’s architecture incorporates security features like secure enclaves, the ultimate responsibility for architecting secure solutions and protecting user data rests with the individual designers utilizing ARM’s technology. This highlights a shared responsibility.
Mr. Abbey envisions an “agentic future” within a few years, where AI systems reason, make complex decisions autonomously, and operate 24/7. This includes autonomous robots, proactive health insights from personal devices, and AI assisting with personalized holiday planning, making lives more productive and efficient.

