AI's growth is straining chip and data center materials

As AI computing demands grow, semiconductors and data centers are hitting physical limits in performance, heat management, electrical efficiency, and reliability. Industry voices argue that the materials used to build this infrastructure now matter as much as the software and models running on it.

Mike Finelli, an executive at chemicals company Syensqo, says advanced materials will be critical to keeping AI infrastructure efficient and sustainable as demand scales. The piece frames materials science as an emerging bottleneck for the AI industry, alongside chips and power.