i卡會員

Demand for AI computing power is rising rapidly and reshaping the future of the global semiconductor industry. As generative AI evolves from passively responding to prompts into “agentic AI” that can autonomously carry out tasks across multiple systems, the amount of computation required for each task is increasing substantially, fueling continued expansion in data centers and AI infrastructure. This transformation involves more than sustained market growth; it represents a fundamental shift in chip architecture—from general-purpose designs to specialization and system-level co-design. At the same time, AI is moving beyond the cloud and into the physical world.