Session Details

[S2]Photonic Computing Solving the Bottleneck of Von-Neumann Computing

Tue. Jul 1, 2025 11:00 AM - 12:30 PM JST
Tue. Jul 1, 2025 2:00 AM - 3:30 AM UTC
Room A (1F Conference Hall)
Organizers: Ryan Michael Hamerly (NTT Research), Mitsumasa Nakajima (NTT Corporation), Nathan Youngblood (University of Pittsburgh)
Presider: Mitsumasa Nakajima (NTT Corporation)
Rapid advancement of machine learning technologies raised critical issues regarding the energy consumption for computations, which motivated various researches on alternative analog computing hardware. Photonic computing, with its unique advantages such as ultra-wide bandwidth and space/wavelength parallelism, offers a promising avenue for dramatically improving computational efficiency. Furthermore, recent breakthroughs in nanophotonics and optoelectronic integration suggest the possibility of large-scale integration of photonic computing engines. This symposium will explore the current state of optical computing integration, algorithms, and applications, discussing both its strengths and weaknesses, as well as its future prospects.

[S2-4]Hyperspectral In-Memory Computing Using Frequency Comb and Programmable Optical Memories

○ByoungJun Park, Mostafa Honari Latifpour, Yoshihisa Yamamoto, Myoung-Gyun Suh (NTT Research)

[S2-5]High-Dimensional Spatially Programmable Photonics

○Martin M Stein1,2,3, Tatsuhiro Onodera1,2, Ryotatsu Yanagimoto1,2, Logan Wright1,2,3, Peter L. McMahon2 (1NTT Research, 2Cornell Univ., 3Yale Univ.)

[S2-6]Scalable Optical Metasurface for On-Edge Visual Intelligence

○Chaoran Huang, Mingcheng Luo, Jiayong Peng, Chester Shu (The Chinese Univ. of Hong Kong)