Sunnyvale, Calif. – July 28, 2021 – NTT Research, Inc., a division of NTT (TYO:9432), today announced that it has entered a joint research agreement with The University of Tokyo’s International Research Center for Neurointelligence (IRCN) to develop Coherent Ising Machine (CIM)-related technologies. The agreement calls for the two research organizations to develop new numerical tools and a simulator for the CIM, an information processing platform based on photonic oscillator networks. The principal investigator (PI) for the three-and-a-half year research project is IRCN Deputy Director Kazuyuki Aihara, a University Professor at the University of Tokyo and expert in the mathematical modeling of complex systems and applications to neurointelligence. His counterpart at NTT Research is Physics & Informatics (PHI) Lab Senior Research Scientist Dr. Satoshi Kako, whose research is focused on the potential capability and application of coherent network computing.
A key component of the PHI Lab’s research agenda, a CIM addresses problems that have been mapped to an Ising model, which is a mathematical abstraction of magnetic systems composed of interacting spins, or angular momentums, of fundamental particles. A primary goal of this joint research with the IRCN is to develop a novel neuromorphic computing principle for combinatorial optimization and machine learning. Combinatorial optimization problems, which a CIM is programmed to solve, require finding an optimal combination of variables from a larger set under various constraints. This project is directed toward finding a new computing principle and algorithms that can be implemented on a modern digital CIM platform. A near-term goal is to provide a field-programmable gate array (FPGA)-based CIM simulator with 16,000 spins and all-to-all couplings.
“Our goal is always to create new scientific knowledge,” said PHI Lab Director Yoshihisa Yamamoto. “We also anticipate that the CIM simulator and digital algorithms that result from this search for knowledge will be used by our numerous collaborators in other research and academic organizations, which is also likely to accelerate the search for applications in this field.”
The NTT Research PHI Lab has embraced an expansive mission of re-thinking computation according to fundamental principles of quantum physics and brain science. Dr. Yamamoto, along with IRCN Project Associate Professor Timothée Leleu and Stanford University professors Surya Ganguli and Hideo Mabuchi, elaborated on this interdisciplinary research agenda, which arguably constitutes a new field of study, in a cover article last year in Applied Physics Letters (APL) titled, “Coherent Ising Machines: Quantum optics and neural network perspectives.” A core part of the IRCN project is previewed in a presentation that Dr. Leleu delivered at the NTT Research Upgrade 2020 summit, titled, “Neuromorphic in Silico Simulator for the CIM.”
Professor Aihara, IRCN Deputy Director and PI for this project, has studied mathematical theory for modeling complex systems and developing trans-disciplinary applications in science and technology. He has developed a theoretical platform composed of advanced control theory of complex systems, complex network theory and nonlinear data analysis and data-driven modeling. On the applications side, he has also worked to bridge biological and clinical studies with human disease prediction and next-generation artificial intelligence (AI).
“This joint research project with NTT Research is an exciting opportunity to continue exploring the intersection of advanced theory and future applications,” Professor Aihara said. “We have a strong foundation and considerable momentum going in and will continue to draw inspiration from advances in mathematical and chaos engineering, optics and neuroscience as the research collaboration unfolds.”
As part of its goal to radically redesign artificial computers, both classical and quantum, the NTT Research PHI Lab has established similar relationships with eight universities. In addition to the IRCN at The University of Tokyo, the PHI Lab has entered joint research agreements with California Institute of Technology (CalTech), Cornell University, Massachusetts Institute of Technology (MIT), Notre Dame University, Stanford University, Swinburne University of Technology, the University of Michigan and the Tokyo Institute of Technology. It is also conducting joint research with the NASA Ames Research Center in Silicon Valley and 1QBit, a private quantum computing software company.
This page is modified and reprinted from an NTT Research press release. The original press release is here.
