Abstract
We study dispersion models and design variations for programmable MEMS photonic networks to analyze scalability of parallel matrix-vector multiplication, which is a core element of commercially viable and energy-efficient photonic neural network accelerator chips.
© 2021 The Author(s)
PDF ArticleMore Like This
Sunil Pai, Nathnael Abebe, Michael Dubrovsky, Rebecca L Hwang, Maxim Karpov, Bogdan Penkovsky, David A. B. Miller, and Olav Solgaard
JTu3A.124 CLEO: Applications and Technology (CLEO:A&T) 2021
Wim Bogaerts, Alain Yuji Takabayashi, Pierre Edinger, Iman Zand, Gaehun Jo, Hamed Sattari, Peter Verheyen, Moises A. Jezzini, Cleitus Antony, Giuseppe Talli, Mehrdad Saei, Saurav Kumar, Cristina Lerma Arce, Marco Garcia Porcel, Niels Quack, Kristinn B. Gylfason, Frank Niklaus, and Umar Khan
IM2A.1 Integrated Photonics Research, Silicon and Nanophotonics (IPR) 2021
Hiromu Toba, Kazuhiro Oda, and Kyo Inoue
SaB1 Optical Amplifiers and Their Applications (OAA) 1995