AI semiconductors are a promising field with significant growth potential and the ability to lead the global market. The AI SoC Research Division is researching, designing, and developing ultra-large artificial neural network SW/HW semiconductors that can be trained based on the Transformer model, which has reached the pinnacle of digital neural networks in achieving human-like intelligence. The goal of our division is to secure the foundational technology for petascale performance gigabyte-level memory-integrated NM-PIM processor semiconductors, and are conducting research for that purpose. Furthermore, to overcome the limitations of traditional von Neumann computing, we are developing neuromorphic computing technologies that simulate neural networks with memory and processing capabilities fused together, enabling ultra-low power and high-performance parallel computing. We are also pursuing the advancement and industrialization of ultra-low power RISC-V edge processors and semiconductor technologies for sensing biological, object, and spatial detection, while having interest in the foundational technologies of controlling quantum computing.