Thursday, April 2, 2026
4:00 pm-5:00 pm
Functional Metal Oxide Thin Films for Bio-Inspired Computing
Recent surge in the development of artificial intelligence (AI) such as ChatGPT has increased the energy demand that outpaces current global energy production capacity. It has also become increasingly clear that existing computing architectures, which rely on the continued scaling of CMOS transistors to improve speed and energy efficiency, are approaching fundamental physical and thermodynamic limits. These architectures are inherently inefficient for data intensive AI systems where the constant shuttling of data between processors and memory, so-called von Neumann bottleneck, dominates energy consumption and latency. New computing platforms such as bio-inspired neuromorphic and quantum computing are at the forefront of research and development to address these challenges. In this talk, I will discuss our recent efforts at CINT to tackle these issues by exploring novel materials and devices for next-generation computing platforms. In particular, we investigate metal oxide thin films to develop bio-inspired devices that can emulate synaptic and neural functionalities of brain. Metal oxide exhibiting ionic transport, phase change, and ferroelectricity enables such devices for memory and information processing in the same physical unit. This co-located storage and computation eliminate the data movement, leading to substantial energy savings and faster operation. Moreover, their intrinsic analog and stochastic properties can be harnessed for in-memory, and analog computing that are naturally suited for modern AI. By moving beyond purely charge-based CMOS toward materials and devices that exploit materials’ physics to emulate analog functions, this approach opens pathways to energy-efficient and bio-inspired computing architectures.
