Semiconductor industry shifts toward AI-driven design and 2nm production

The semiconductor industry is seeing significant strategic shifts as companies prepare for the rise of physical AI and advanced computing. Cadence is implementing a ‘three-layer cake’ strategy to expand its reach from traditional chip design EDA to full system design simulation. This approach utilizes hardware acceleration, high-performance computational software, and AI-driven design workflows to address the complexities of 3DIC and thermal management. Cadence aims to provide the essential computational tools—described as ‘shovels’ in an AI-driven market—to help clients manage rapid iterations in sectors like automotive and robotics.

In Japan, Rapidus is advancing its plans to mass-produce 2nm semiconductors, targeting the high demand from AI and data center markets. Rapidus CEO Junji Koike stated that the company is on track with its production schedule and intends to compete through speed, offering significantly shorter lead times for prototypes. Comparing its value proposition to the premium pricing of a Shinkansen bullet train, Koike noted that customers may be willing to pay more for rapid delivery to bridge the global supply gap. The company is currently in negotiations with dozens of potential clients, including major tech firms, and is working closely with IBM to refine its technology.