IEEE Circuits and Systems Magazine - Q2 2022 - 27

upon the proposed circuit. The network demonstrated
capability of pattern recognition in the optical domain.
[330] implemented a neuromorphic photonic network
to solve an ordinary differential equation system called
a Lorenz attractor, and it achieved 294 × speedup compared
to a CPU baseline. Though photonic neuromorphic
computing is still far from practical, it has the potential to
exceed electronic devices' performance by many orders
of magnitude [337].
While the neuromorphic systems implemented using
aforementioned emerging device technologies have
demonstrated great potentials, they also face significant
challenges. How to improve their scalability, flexibility
and reliability will continue to be the research direction
in the future.
VI. Conclusions
As a bio-inspired computing paradigm, neuromorphic
computing has great potentials in accelerating computational
neuroscience, and enabling energy efficient
solutions for machine intelligence. Due to its unique
way of encoding and processing information, it is also
believed to be particularly promising for sensor and
control-based applications that interact with the physical
environment. In this survey, we reviewed different
computation models, learning algorithms, information
coding schemes, and hardware architectures of neuromorphic
computing. With more and more research
efforts in academia and industry, we anticipate that
breakthroughs in more reliable learning algorithms
and more efficient implementations will be seen in the
near future.
Acknowledgment
This work is partially supported by the National Science
Foundation I/UCRC ASIC (Alternative Sustainable and Intelligent
Computing) Center (CNS-1822165).
The work was received and approved for public release
by AFRL on August 24th, 2021, case number AFRL2021-2820.
Any Opinions, findings, and conclusions or
recommendations expressed in this material are those
of the authors and do not necessarily reflect the views
of AFRL or its contractors.
Amar Shrestha received the B.S. degree
in electrical and electronics engineering
from the Department of Electrical
and Electronics Engineering in
Kathmandu University, Dhulikhel, Nepal,
in 2013. He has completed the Ph.D. degree
from the Department of Electrical Engineering
and Computer Science, Syracuse University, Syracuse,
NY, USA with a disseration titled " Inference and
SECOND QUARTER 2022
Zaidao Mei was born in China. He received
the B.S. degree from the Ohio
State University, Columbus, USA, 2017,
and the master's degree from Syracuse
University, Syracuse, NY, USA. He is currently
pursuing the Ph.D. degree in
computer engineering with the Department of EECS,
Syracuse University. His current research interests include
machine learning and neuromorphic computing.
Daniel Patrick Rider received his B.S.
in Mathematics and Electrical Engineering
from University at Buffalo, Buffalo,
NY (2012), and his M.S. in Electrical Engineering
from Syracuse University,
Syracuse, NY (2020). He is currently
pursuing a Ph.D. in Electrical and Computer Engineering
at Syracuse University. His research interests include
neuromorphic computing, simultaneous localization
and mapping, and fault tolerance and reliability in computing
devices.
Qing Wu, a member of the scientific
and professional cadre of senior executives,
is the Senior Scientist for Processing
and Exploitation, Information
Directorate, Air Force Research Laboratory,
Rome, New York. He serves the
Air Force as the principal scientific authority and independent
researcher in the research, development, adaptation
and application of technologies to manage,
process and exploit massive amounts of multi-intelligence,
multi-domain data. Dr. Wu's primary objectives
are to advance knowledge of processing and exploitation
means, including emerging technologies, to yield
decision-quality information derived from diverse
IEEE CIRCUITS AND SYSTEMS MAGAZINE
27
Learning in Spiking Neural Networks for Neuromorphic
Systems. " His main research interests include
neuromorphic computing, deep learning and natural
language understanding with publications in many
conferences including DAC, ICCAD, IJCNN, etc.
He is currently a Research Scientist at Amazon in
Seattle, WA.
Haowen Fang received the M.S. and
Ph.D. degrees from the Department of
Electrical Engineering and Computer
Science, Syracuse University, Syracuse,
NY, USA, in 2018 and 2021, respectively.
His research interests include deep
learning, neuromorphic computing, neural
network acceleration, and FPGA.
IEEE Circuits and Systems Magazine - Q2 2022

Table of Contents for the Digital Edition of IEEE Circuits and Systems Magazine - Q2 2022
