Efficient and flexible edge computing

Flashing Technology Team Concept

Physical reservoir computing can be used to perform high-speed processing for artificial intelligence with low power consumption.

Researchers from Japan design a tunable physical reservoir device based on dielectric relaxation at an electrode-ionic liquid interface.

In the near future, more and more AI processing will need to be done at the edge, close to the user and where the data is collected rather than on a distant computer server. This will require high-speed data processing with low power consumption. Physical reservoir computing is an attractive platform for this purpose, and a new breakthrough from scientists in Japan has just made this much more flexible and practical.

Physical Reservoir Computing (PRC), which is based on the transient response of physical systems, is an attractive machine learning framework that can perform high-speed processing of time-series signals at low power. However, PRC systems have low adjustability, which limits the signals it can process. Now researchers in Japan present ionic liquids as an easily tunable physical reservoir device that can be optimized to process signals over a wide range of time scales simply by changing its viscosity.

Artificial intelligence (AI) is rapidly becoming ubiquitous in modern society and will see wider implementation in the coming years. In applications involving sensors and Internet of Things devices, the norm is often edge AI, a technology in which computation and analytics are performed close to the user (where data is collected) and not far away in the distance. a centralized server. This is because edge AI has low power requirements as well as high-speed data processing capabilities, characteristics that are particularly desirable in real-time time-series data processing.

Time scale of signals commonly produced in living environments

Time scale of signals commonly produced in living environments. The response time of the ionic liquid PRC system developed by the team can be tuned to optimize the processing of such real-world signals. Credit: Kentaro Kinoshita of TUS

In this sense, physical reservoir computing (PRC), which is based on the transient dynamics of physical systems, can greatly simplify the edge AI computing paradigm. This is because PRC can be used to store and process analog signals into ones that the edge AI can efficiently work with and analyze. However, the dynamics of solid PRC systems are characterized by specific time scales that cannot be easily tuned and are typically too fast for most physical signals. This mismatch in time scales and its poor controllability make PRC largely unsuitable for real-time signal processing in living environments.

To address this problem, a research team from Japan consisting of Professor Kentaro Kinoshita and Sang-Gyu Koh, a doctoral student at Tokyo University of Science, and principal investigators Dr. Hiroyuki Akinaga, Dr. Hisashi Shima, and Dr. Yasuhisa Naitoh of the National Institute of Advanced Industrial Science and Technology, proposed, in a new study published in the journal scientific reports, the use of liquid PRC systems instead. “Replacing conventional solid reservoirs with liquids should lead to AI devices that can learn directly on the timescales of environmentally generated signals, such as voice and vibrations, in real time,” explains Professor Kinoshita. “Ionic liquids are stable molten salts that are entirely composed of freely moving electrical charges. The dielectric relaxation of the ionic liquid, or how its charges rearrange in response to an electrical signal, could be used as a reservoir and holds great promise for edge AI physical computing.”

Computation of ionic liquid-based reservoirs

The response of the ionic liquid PRC system can be tuned to optimize processing of a wide range of signals by changing their viscosity by adjusting the length of the cationic side chain. Credit: Kentaro Kinoshita of TUS

In their study, the team designed a PRC system with an ionic liquid (IL) of an organic salt, 1-alkyl-3-methylimidazolium bis(trifluoromethane sulfonyl)imide ([Rmim+] [TFSI] R = ethyl (e), butyl (b), hexyl (h) and octyl (o)), whose cationic part (the positively charged ion) can be easily varied with the length of a chosen alkyl chain. They fabricated gold gap electrodes and filled the gaps with the IL. “We found that the time scale of the reservoir, although complex in nature, can be directly controlled by the viscosity of the IL, which is dependent on the length of the cationic alkyl chain. Changing the alkyl group in organic salts is easy to do and presents us with a controllable and designable system for a range of signal lifetimes, allowing for a wide range of future computational applications,” says Professor Kinoshita. By tuning the alkyl chain length between 2 and 8 units, the researchers achieved characteristic response times ranging from 1 to 20 µs, with longer alkyl side chains leading to longer response times and higher learning performance. of adjustable AI of the devices.

The system’s tunability was demonstrated using an AI image identification task. The AI ​​was presented with a handwritten image as input, which was represented by rectangular pulse voltages 1 µs wide. By increasing the length of the sidechain, the team made the transient dynamics approach that of the target signal, and the rate of discrimination improved for higher chain lengths. This is because, compared to [emim+] [TFSI]in which the current relaxed to its value in about 1 µs, the IL with a longer side chain and in turn longer relaxation time better retained the history of the time series data, improving identification[{” attribute=””>accuracy. When the longest sidechain of 8 units was used, the discrimination rate reached a peak value of 90.2%.

Input Signal Conversion Through Ionic Liquid Based PRC System

Input signal conversion through the ionic liquid-based PRC system. The reservoir output in the form of current response (top and middle) to an input voltage pulse signal (bottom) are shown. If the current decay (dielectric relaxation) is too fast/slow, it reaches its saturation value before the next signal input and no history of the previous signal is retained (middle image). Whereas, if the current response attenuates with a relaxation time that is properly matched with the time scales of the input pulse, the history of the previous input signal is retained (top image). Credit: Kentaro Kinoshita from TUS

These findings are encouraging as they clearly show that the proposed PRC system based on the dielectric relaxation at an electrode-ionic liquid interface can be suitably tuned according to the input signals by simply changing the IL’s viscosity. This could pave the way for edge AI devices that can accurately learn the various signals produced in the living environment in real time.

Computing has never been more flexible!

Reference: “Reservoir computing with dielectric relaxation at an electrode–ionic liquid interface” by Sang-Gyu Koh, Hisashi Shima, Yasuhisa Naitoh, Hiroyuki Akinaga and Kentaro Kinoshita, 28 April 2022, Scientific Reports.
DOI: 10.1038/s41598-022-10152-9

Kinoshita Kentaro is a Professor at the Department of Applied Physics at Tokyo University of Science, Japan. His area of interest is device physics, with a focus on memory devices, AI devices, and functional materials. He has published 105 papers with over 1600 citations to his credit and holds a patent to his name.

This study was partly supported by JSPS KAKENHI Grant Number JP20J12046.

Tokyo University of Science (TUS) is a well-known and respected university, and the largest science-specialized private research university in Japan, with four campuses in central Tokyo and its suburbs and in Hokkaido. Established in 1881, the university has continually contributed to Japan’s development in science through inculcating the love for science in researchers, technicians, and educators.

Leave a Reply

Your email address will not be published.