Advances in artificial neural networks and Big Data analytics have begun to deliver impressive learning, reasoning and human-machine interaction capabilities on computing hardware founded on silicon CMOS technology. However, the current approach to solve such cognitive problems on conventional computers is universally considered as a time-consuming task which requires datacenter-scale computational resources. Therefore, there has been increasing interest in the field of neuromorphic engineering and brain-inspired computing to build such networks on native hardware, with an expectation of achieving a brain-like efficiency and large performance gain as well as new functionalities. This requires a systematic approach combining the expertise of material scientists to study, refine and develop neuromorphic materials with architecture-level understanding.
The goal of this symposium is to provide a forum to unite researchers who are engaged in the study of material research across neuromorphic computing technologies. This symposium will cover the scientific and technological exploration and advances of the development, characterization and system-level integration of new materials and devices for a variety of neuromorphic applications. The materials of interest include: resistive switching materials, conductive filament materials, phase-change, ferroelectric and tunneling-based materials. Theoretical and modeling studies to define clear performance metrics, or investigating material and device requirements for specific algorithms and architectures, are also welcome. Invited talks will attempt to bridge the gap between interdisciplinary topics such as material science/engineering, neuroscience, computer science, algorithms, systems and architecture-level considerations in order to accelerate the discussion and development of these materials toward practical applications.