Breakthrough in Neuromorphic Computing: University of Tokyo Researchers Unveil Brain-Inspired Device Using Ferroelectric Oxides
Tokyo, Japan – September 9, 2025 – The University of Tokyo has announced a significant advancement in the field of neuromorphic computing with the successful demonstration of a brain-like device that leverages the colossal resistance change phenomenon in ferroelectric oxides. Published today, September 9, 2025, at 05:00 JST, this groundbreaking research paves the way for highly energy-efficient and powerful artificial intelligence systems.
The research, detailed in a press release by the University of Tokyo, centers on the innovative use of ferroelectric oxide materials. These unique materials exhibit a spontaneous electric polarization that can be switched by an external electric field. Crucially, this switching process in certain ferroelectric oxides is accompanied by a dramatic change in their electrical resistance, a phenomenon known as colossal resistance change.
Inspired by the fundamental workings of the human brain, which relies on the complex interplay of neurons and synapses to process information, the researchers have engineered a device that mimics the behavior of these biological components. Synapses, the junctions between neurons, are responsible for modulating the strength of signals transmitted between them. This synaptic plasticity is key to learning and memory.
The newly developed device effectively replicates the function of a synapse by utilizing the colossal resistance change within the ferroelectric oxide. The resistance of the device can be precisely controlled by applying electrical pulses, analogous to how synaptic strength is modified in the brain. This allows the device to “learn” and “remember” by adjusting its conductivity in response to input signals.
The implications of this breakthrough are profound. Current AI hardware, while powerful, is often energy-intensive. Neuromorphic computing aims to create hardware that operates more like the brain, promising vastly improved energy efficiency and processing capabilities for tasks such as pattern recognition, machine learning, and artificial intelligence.
The use of ferroelectric oxides offers several advantages for neuromorphic applications. These materials are known for their stability and their ability to store information for extended periods. Furthermore, the colossal resistance change allows for a significant dynamic range in conductivity, which is essential for replicating the nuanced signaling observed in biological neural networks.
This research represents a significant step forward in the quest to build truly brain-inspired computing systems. By harnessing the remarkable properties of ferroelectric oxides, the University of Tokyo team has demonstrated a novel and promising approach to creating the next generation of AI hardware. The potential for more efficient, powerful, and biologically plausible artificial intelligence systems has just taken a significant leap forward.
AI has delivered the news.
The answer to the following question is obtained from Google Gemini.
東京大学 published ‘強誘電体酸化物の巨大抵抗変化を利用して脳型素子を実現’ at 2025-09-09 05:00. Please write a detailed article about this news in a polite tone with relevant information. Please reply in English with the article only.