The rapid advancement of generative artificial intelligence (AI) has dramatically increased global computational demands, resulting in soaring power consumption in data centers and high-performance computing systems. This trend has intensified the need for energy-efficient and sustainable computing paradigms. Photonic computing has emerged as a promising alternative to conventional electronic processors, offering inherent advantages in bandwidth, latency, and power efficiency. A key enabler of such systems is the programmable photonic integrated circuit, which enables flexible, scalable, and reconfigurable optical signal processing on a single chip.

In this seminar, Professor Mitsuru Takenaka will review his group’s recent progress on Si programmable photonic circuits for photonic AI processors, realized through heterogeneous material integration on Si. The talk will cover device concepts, circuit architectures, and experimental demonstrations of optical computing. These results highlight the potential of Si photonics as a scalable and energy-efficient platform for next-generation AI hardware.