Date of Award

8-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Computer Science

Major Professor

James S. Plank

Committee Members

Douglas S. Aaron, Mark E. Dean, Garrett S. Rose

Abstract

Moore’s Law fairly accurately modelled advancements in traditional computing architectures for multiple decades, but it has come to an end. This has led researchers to put more focus on alternative computing architectures such as neuromorphic computing. DANNA (Dynamic Adaptive Neural Network Array) is a computing architecture that was designed in 2014 to meld features of recurrent, spiking, plastic neuromorphic computing systems with very efficient hardware implementations. Its hardware design and FPGA implementation preceded any software support or simulation. This work describes the software support for DANNA, including four different simulators, that has enabled TennLAB to explore the capabilities of the architectures. Additionally, we generalized a well-known neuromorphic experiment from 2008 to fit within the TennLAB software structure. We use this experiment to compare the capabilities of DANNA and TennLAB’s other neuromorphic models.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS