优化式AI

Search documents
EDA巨头高管:三维集成电路的未来,不仅是物理结构堆叠
Guan Cha Zhe Wang· 2025-09-03 05:40
Core Insights - The semiconductor industry is experiencing a significant transformation driven by advancements in AI, particularly in chip design and system development [1][2][3] - Cadence's vision for the future of chip design involves the integration of Agentic AI to automate the design process, moving away from manual coding and layout [1][4] - The market forecast for the semiconductor industry has been revised upwards, with expectations to exceed $1.2 trillion by 2030, largely due to the growth in AI computing and data centers [2] Group 1: AI and Chip Design - The emergence of generative AI and Agentic AI will lead to automated chip design based on high-level functional descriptions, enhancing user experience [1][2] - Advanced packaging technologies, such as 3D integrated circuits, are essential for overcoming performance bottlenecks in complex semiconductor designs [1][2] - Cadence's "three layer cake" concept emphasizes the need for collaboration between design tools, IP development, and hardware solutions to meet dynamic customer requirements [2] Group 2: EDA and AI Integration - The traditional methods of simulation and AI in EDA are evolving, with AI now capable of addressing complex physical modeling and automated design challenges [3][4] - Cadence has integrated "Optimization AI" into over 50% of its tools, with plans to exceed 80% in the next two years, enhancing speed, quality, and error detection [4] - The introduction of the JedAI platform allows users to interact with tools using natural language, facilitating a shift towards a "dialogue-based interaction" era in EDA [3][4] Group 3: Future Vision - The long-term vision for design automation involves users simply inputting functional requirements, with Agentic AI generating netlists and executing design processes autonomously [4] - Cadence aims to be a partner in intelligent system design, extending capabilities from chips to packaging and circuit boards [4]