The Design of Neuromorphic Controller System Built with Memristor Crossbars

In the development history of microprocessors, although computing efficiency of solid state circuits keeps improving by following technology scaling, the data transportation efficiency between CPU cores and storage systems continue drifting down and starts dominating the energy consumption of the entire system. This phenomenon is referred to as the ?memory wall?. The neuromorphic computing inspired by the working mechanism of human brains effectively reduces the data communication cost and consequently, achieves very high computation efficiency. On the one hand, since both the frequency of the spikes and their relative timing are carrying on the transmitted information, the spikes can be very short and sparse to minimize the amount of the relocated electrical charge; On the other hand, neuromorphic computing also minimizes the data communication distance by distributing the data into the memories (i.e., synapses) close to the associated computing units (i.e., neurons) throughout the entire system. However, neuromorphic systems, such as cortical processor, require very high connectivity and flexible reconfigurability, which commonly consumes a large volume of memory and computing resources, incurring high design complexity and hardware cost in conventional CMOS implementation.
The objective of the project is to investigate the neuromorphic computing systems built with the emerging memristor technology. We plan to design a neuromorphic computing controller at IBM 130nm process which can be integrated with real memristor crossbar array(s) at board level. To represent the use of the bio-inspired computing system in extreme power-efficient neuromorphic applications and high-performance computation acceleration, we will implement both spike-based and level-based controllers.

Senior Personnel:	Prof. Hai (Helen) Li (PI), Prof. Yiran Chen (co-PI)
Sponsors:	Air Force Research Lab (AFRL)
Team Members:	Chenchen Liu, Bonan Yan, Zheng Li, Chaofei Yang, Yandan Wang
Collaborators:	Prof. Jianhua Yang, University of Massachusetts at Amherst, Prof. Qiangfei, University of Massachusetts at Amherst
Start Date:	2015

Related Publications