Architectural impacts of emerging transistors

Abstract

At present there is much effort to determine if emerging information processing devices could have a positive impact on the performance of non-Boolean/non-von Neumann computer architectures. We explore this topic here by specifically considering how emerging transistor technologies might impact cellular neural networks (CNNs). For CNNs, prior work suggests that new transistor structures could be employed to better facilitate non-binary outputs -that in turn reduce the number of template/programming operations as well as the hardware paths needed to solve a given problem. Here, we present analysis that considers how the above approach could impact the performance/energy of the cells that comprise the CNN. We also consider how characteristics of other emerging transistor technologies could positively impact CNNs -particularly with respect to reduced program complexity.

Understanding the landscape of accelerators for vision

“…However, non-linear templates often reduce the number of programming steps required to solve a particular problem, when compared to an algorithm that employs only linear templates. Initial work [44] suggests that non-linearities associated with SymFETs can be used to efficiently realize non-linear templates. As an example, to perform thresholding on an interval using a band pass filter, three sequential linear operations are needed.…”

TFET-based Operational Transconductance Amplifier Design for CNN Systems

“…Previous research [9,10,11,12] have shown positive results by utilizing emerging devices in multiple CNN contexts. Specifically, [9] shows TFETs can be utilized as nonlinear resistive element in CNNs to improve power efficiency without sacrificing performance.…”