PNNL

Abstract

Artificial Intelligence (AI) has a hardware problem. Progress in AI, like progress in other scientific fields, is accelerated by tightly coupling users, designers, and manufacturers into a cycle of innovation that provides clear demand feedback. Most of the conversation about AI’s hardware problem focuses on technical obstacles and solutions. However, the AI hardware problem is also market-driven: the global semiconductor industry, which creates the hardware on which all AI operates, faces cyclical and structural challenges that pose potential roadblocks to ongoing innovation by AI designers and users. The most celebrated AI chips today all share one thing in common: they are made by Taiwan Semiconductor Manufacturing Corporation. This consolidation of leading edge semiconductor manufacturing in one company in one country poses two problems: a short-term strategic risk and a long-term innovation risk. Both problems remain under-explored because advances in AI algorithms in the past decade have compensated for a relative lack of advances in AI hardware.