The seminar on the new technology route for high-performance AI substrate materials was held in Beijing.

According to the China Electronic Information Industry Development Research Institute, on the afternoon of August 14, the AI High-Performance New Substrate Technology Route Seminar jointly organized by the Cyber Development Institute's Artificial Intelligence Scene Innovation and Smart Chip Evaluation Key Laboratory and the Artificial Intelligence Industry Committee was held at the China Center for Information Industry Development in Beijing. The conference focused on the engineering feasibility and industrialization prospects of new substrate technology routes against the backdrop of explosive growth in AI computing power. It invited experts from fields such as material research and development, equipment manufacturing, packaging technology, chip design, and intelligent computing servers to jointly assess technology trends and discuss paths for industrial collaboration. The participating experts unanimously agreed that new substrates are currently at an industrialization window period, with the main bottlenecks being insufficient collaboration among "materialsequipmentprocessesend products," lack of standards, and absence of pilot testing platforms. Key materials, specialized equipment, and advanced packaging could form new bottlenecks due to supply-demand mismatches and concentration in the supply chain; the lack of reliability data and unified testing standards has also led to "material companies being hesitant to invest and end product companies being reluctant to adopt." Experts suggested enhancing capacity monitoring and early warning systems, promoting diversified supply and technological backups, establishing cross-domain joint problem-solving mechanisms, and building demonstration application pilot zones to accelerate localization through scenario-driven approaches, thereby facilitating the industrial closed loop from laboratory to production line, and from materials to end products.