IEEE Symposium on Memristor and Memristor-based Computing Systems (IEEE MMCS)
The recent slowdown of semiconductor technology scaling and the prominent von Neumann bottleneck have begun to motivate increasing investment in neuromorphic computing systems, i.e., system realization of brain-inspired computing architecture. In addition to the various designs of neural circuits and synapse networks being implemented on conventional CMOS platforms, the application of emerging nano-devices in neuromorphic computing system designs, such as memristor devices, are also being actively investigated as a promising approaches to achieve ultra-high computing capacity and performance. However, there are still many significant technical challenges that limit realization of the relevant research and severely hinder upscaling of neuromorphic computing system toward the complexity needed by real-world applications.
This Symposium is devoted to several fundamental challenges in the research and development of memristor-based neuromorphic computing system designs via: 1) reliably and efficiently perform neuromorphic computation on memristor crossbars, taking into account current ideas surrounding spiking-based computation, computing workload partitioning, and etc.; 2) modeling methodologies, as well as the relevant optimizations, in order to accurately simulate the electrical properties of memristor devices, and reduce the computing complexity of memristor-based neuromorphic computing system while enhancing training reliability; and 3) physical and architecture design techniques for robust and scalable memristor-based neuromorphic computing system design.
It is aiming to publish the frontiers of theories, modeling methods, and design techniques for memristor crossbar based neuromorphic computing and benefit both the computer system and circuit communities through a truly interdisciplinary scalable design flow for neuromorphic computing system built on emerging memristor technology.
The list of possible topics includes, but is not limited to:
- Brain-like computing theory and systems
- Memristor materials, devices and circuits
- Memristor-based neural networks
- Memristor-based machine learning
- Memristor-based neuromorphic computing systems
- Modeling methodologies of memristor/memristor-based crossbars
- Optimization methods to reduce the computing complexity
- Efficient training analysis for memristor-based deep learning
- Novel memristor-based network architecture based on theoretical analysis
- Neurodynamics theory and its applications of memristor-based systems