Invited - Temporal information processing for in-sensor computing based on amorphous IGZO phototransistor
Conference Dates
May 15-18, 2023
Abstract
On facing the massive and unstructured data processing, it is imperative to emulate artificial neural networks with new physical hardware architectures in addition to software-based approaches, to overcome the barrier of the von Neumann bottleneck. By mimicking the human visual sensing system, the optoelectronic devices, which can perform data compression and reduce the network size through the reconstruction of input signals, are promising to develop the neuromorphic in-sensor computing for minimizing the time latency as well as improving the energy efficiency. In this work, we demonstrate an amorphous indium-gallium-zinc-oxide (a-IGZO) phototransistor with ZrOx high-k dielectric layer with distinct responses to various optical stimulation inputs. Due to the persistent photoconductivity (PPC) effect of a-IGZO after lighting, our device is able to exhibit synaptic functions via the application of 405 nm light spikes, such as paired-pulse facilitation (PPF) and short-term memory (STM). Furthermore, in order to perform the temporal optical signals processing, the a-IGZO phototransistor is stimulated by four-timeframe temporal pulse streams composed of 405 nm light spikes and it expresses the different temporal responses. The distinct output photocurrent response reveals that the a-IGZO phototransistor can be applied to distinguish the time-series input light signals. Accordingly, the a-IGZO phototransistor have a promising potential for processing optical temporal information and can possibly be implemented for visual in-sensor computing techniques.
Please click Download on the upper right corner to see the full abstract.
Recommended Citation
Jen- Sue Chen, Ching-Hsiang Yang, and Li-Chung Shih, "Invited - Temporal information processing for in-sensor computing based on amorphous IGZO phototransistor" in "Semiconductor Technology for Ultra Large Scale Integrated Circuits and Thin Film Transistors (ULSIC VS TFT 8)", Y. Kuo, Texas A&M University, USA Eds, ECI Symposium Series, (2022). https://dc.engconfintl.org/ulsic_vs_tft_8/19