Machine Learning Technologies for Semiconductor Manufacturing

As the size of semiconductor devices decreases and the structures and materials become increasingly complex, manufacturing these devices is becoming more difficult. The complexity of IC research and development (R&D) and the scale of high-volume manufacturing (HVM) have greatly increased the cost and time required to reach final production. Chipmakers, equipment manufacturers and software companies are exploring and deploying machine learning (ML) technology in a wide range of applications, including process development, production maintenance, metrology, and yield improvement, to address these scale-up issues. With more than a decade of expertise deploying ML technology for semiconductor manufacturing, Lam Research has developed several smart tools and ML solutions to optimize quality, efficiency, and productivity and accelerate innovation in semiconductor manufacturing. In this paper, two of Lam's smart tools — SEMulator3D® and Equipment Intelligence® Data Analyzer (EI-DA) —are introduced to demonstrate how Lam's advanced technology can be used to efficiently manufacture state-of-the-art microchips during R&D and HVM.