Parallel convolutional processing using an integrated photonic tensor core

Autor: Nathan Youngblood, Anton Lukashchuk, Abu Sebastian, Maxim Karpov, M. Le Gallo, Tobias J. Kippenberg, Wolfram H. P. Pernice, Johannes Feldmann, Harish Bhaskaran, C.D. Wright, Arslan S. Raja, Maik Stappers, Helge Gehring, Junqiu Liu, Xin Fu, Xuan Li
Rok vydání: 2021
Předmět:
FOS: Computer and information sciences
Computer science
Computer Science - Emerging Technologies
Physics::Optics
FOS: Physical sciences
Information technology
02 engineering and technology
Integrated circuit
01 natural sciences
7. Clean energy
law.invention
Application-specific integrated circuit
law
0103 physical sciences
Electronic engineering
Frequency combs
010302 applied physics
Nanophotonics and plasmonics
Multidisciplinary
business.industry
Bandwidth (signal processing)
Disordered Systems and Neural Networks (cond-mat.dis-nn)
Condensed Matter - Disordered Systems and Neural Networks
021001 nanoscience & nanotechnology
Emerging Technologies (cs.ET)
CMOS
Transmission (telecommunications)
visual_art
Electronic component
visual_art.visual_art_medium
Hardware acceleration
Photonics
0210 nano-technology
business
Physics - Optics
Optics (physics.optics)
Zdroj: Nature
ISSN: 1476-4687
0028-0836
Popis: With the proliferation of ultrahigh-speed mobile networks and internet-connected devices, along with the rise of artificial intelligence (AI)(1), the world is generating exponentially increasing amounts of data that need to be processed in a fast and efficient way. Highly parallelized, fast and scalable hardware is therefore becoming progressively more important(2). Here we demonstrate a computationally specific integrated photonic hardware accelerator (tensor core) that is capable of operating at speeds of trillions of multiply-accumulate operations per second (10(12) MAC operations per second or tera-MACs per second). The tensor core can be considered as the optical analogue of an application-specific integrated circuit (ASIC). It achieves parallelized photonic in-memory computing using phase-change-material memory arrays and photonic chip-based optical frequency combs (soliton microcombs(3)). The computation is reduced to measuring the optical transmission of reconfigurable and non-resonant passive components and can operate at a bandwidth exceeding 14 gigahertz, limited only by the speed of the modulators and photodetectors. Given recent advances in hybrid integration of soliton microcombs at microwave line rates(3-5), ultralow-loss silicon nitride waveguides(6,7), and high-speed on-chip detectors and modulators, our approach provides a path towards full complementary metal-oxide-semiconductor (CMOS) wafer-scale integration of the photonic tensor core. Although we focus on convolutional processing, more generally our results indicate the potential of integrated photonics for parallel, fast, and efficient computational hardware in data-heavy AI applications such as autonomous driving, live video processing, and next-generation cloud computing services.
Databáze: OpenAIRE
Externí odkaz: