19 December 2017 in IEEE Journal of Solid-State Circuits
A versatile reconfigurable accelerator architecture for binary/ternary deep neural networks is presented. In-memory neural network processing without any external data accesses, sustained by the symmetry and simplicity of the computation of the binary/ternaty neural network, improves the energy efficiency dramatically. The prototype chip is fabricated, and it achieves 1.4 TOPS (tera operations per second) peak performance with 0.6-W power consumption at 400-MHz clock. The application examination is also conducted.
29 December 2021 in Advanced Science
Nonlinear dynamical systems serving reservoir computing enrich the physical implementation of computing systems. A method for building physical reservoirs from electrochemical reactions is provided, and the potential of chemical dynamics as computing resources is shown. The essence of signal processing in such systems includes various degrees of ionic currents which pass through the solution as well as the electrochemical current detected based on a multiway data acquisition system to achieve switchable and parallel testing. The results show that they have respective advantages in periodic signals and temporal dynamic signals. Polyoxometalate molecule in the solution increases the diversity of the response current and thus improves their abilities to predict periodic signals. Conversely, distilled water exhibits great computing power in solving a second‐order nonlinear problem. It is expected that these results will lead to further exploration of ionic conductance as a nonlinear dynamical system and provide more support for novel devices as computing resources.
12 February 2021 in Physical Review Applied
The potential of nonlinear dynamical systems serving as reservoirs has attracted much attention for the physical realization of reservoir computing (RC). Here, we propose a hardware system working as a reservoir with one simple form of nonlinearity that reflects the intrinsic characteristics of the materials. We show that insufficient dynamics in such physical systems can perform like complex dynamical systems with the assistance of external controls. Based on the idea of spatial multiplexing, this dynamical system is studied under two frameworks. The correlation between structural adjustments of the reservoir and system performance in processing various types of task is proposed. Our results are expected to enable the development of material-based devices for RC.
18 February 2021 in IEEE Transactions on Circuits and Systems for Video Technology
The rising demand for high quality display has ensued active research in high dynamic range (HDR) imaging, which has the potential to replace the standard dynamic range imaging. This is due to HDR’s features like accurate reproducibility of a scene with its entire spectrum of visible lighting and color depth. But this capability comes with expensive capture, display, storage and distribution resource requirements. Also, display of HDR images/video content on an ordinary display device with limited dynamic range requires some form of adaptation. Many adaptation algorithms, widely known as tone mapping (TM) operators, have been studied and proposed in the last few decades. In this article, we present a comprehensive survey of 60 TM algorithms that have been implemented on hardware for acceleration and real-time performance. In this state-of-the-art survey, we will discuss those TM algorithms which have been implemented on GPU, FPGA, and ASIC in terms of their hardware specifications and performance. Output image quality is an important metric for TM algorithms. From our literature survey we found that, various objective quality metrics have been used to demonstrate the quality of those algorithms hardware implementation. We have compiled those metrics used in this survey, and analyzed the relationship between hardware cost, image quality and computational efficiency. Currently, machine learning-based (ML) algorithms have become an important tool to solve many image processing tasks, and this article concludes with a discussion on the future research directions to realize ML-based TM operators on hardware.
30 June 2023 in Advanced Functional Materials
The human brain possesses an exceptional information processing capability owing to the 3D and dense network architecture of numerous neurons and synapses. Brain‐inspired neuromorphic hardware can also benefit from 3D architectures, such as high integration of circuits and acquisition of highly complex dynamical systems. In this study, for future 3D neuromorphic engineering, 3D conductive polymer networks consisting of poly(3,4‐ethylenedioxy‐thiophene) doped with poly(styrene sulfonate) anions (PEDOT:PSS) are successfully and stably fabricated between multiple electrodes from scratch in precursor solution by electropolymerization. The networks efficiently emulate the 3D local connections between neighboring neurons observed in the cortex. This novel technology, which allows 3D conductive wiring only between desired electrodes, is unprecedented and has potential as an underlying technology for 3D integration. Furthermore, the experimental results also conclusively prove that conductance modification can be performed by manipulating the physical and chemical properties of 3D branch‐wired conductive polymer wires, thus demonstrating for the first time the feasibility of neuromorphic wetware with enhanced biological plausibility in the subsequent post‐Moore era.