CIES Consortium
National Projects
NEDO Development of technology for designing energyefficient AI semiconductor chips and systems
Development of design technology to accelerate industrial application of AI edge computing
Design efficiency of AI processing semiconductors by CMOS/spintronicshybrid technology and its applications
-
Prof.
Takahiro Hanyu -
Prof.
Ko Yoshikawa -
Prof.
Shoji Ikeda
The research project on "Design efficiency of AI processing semiconductors by CMOS/spintronics-hybrid technology and its applications (principal investigator: Tohoku University; representative: prof. Takahiro Hanyu)" proposed in 2022 was accepted by the new energy and industrial technology development organization (NEDO). The project was adopted by NEDO for "development of AI chips and nextgeneration computing technology enabling high-efficiency and highspeed processing / development of design technology to accelerate industrial application of AI edge computing".
The current AI processing handles a large amount of data and requires high-speed computation, which results in huge power consumption. In particular, since the allowable power consumption is limited at the edge, innovative and power-efficient AI computing technologies are desired.
This research aims to develop a high power-efficient computing technology suitable for edge AI that maximizes the characteristics of CMOS technology, which is currently the mainstream, and spintronics technology, which is highly compatible, non-volatile, and areaefficient, by integrating these technologies. In addition, as its social implementation, we will apply the technology to in-vehicle systems and to next-generation surveillance systems such as watchdog systems.
