MIXED SIGNAL PROCESSORS

摘要

Various processor architectures for mixed signal computation exploit the unique characteristics of advanced CMOS technologies, such as fin-based, multi-gate field effect transistors, and/or emerging technologies such as tunnel field effect transistors (TFETs). The example processors disclosed herein are cellular neural network (CNN)-inspired and eliminate the need for voltage controlled current sources (VCCSs), which have previously been utilized to realize feedback and feed-forward templates in CNNs and are the dominant source of power consumption in a CNN array. The example processors replace VCCSs with comparators, which can be efficiently realized with TFETs given their high intrinsic gain. Power efficiencies are in the order of 10,000 giga-operations per second per Watt (GOPS/W), which represents an improvement of more than ten times over state-of-the-art architectures seeking to accomplish similar information processing tasks.

主权项

1. A processor including a homogenous array of cells, with each of the cells capable of receiving an analog input, communicating with adjacent cells, and producing a digital output, each of the cells comprising:
a comparator; a first input to the comparator that includes a voltage ramp that is shared by the homogenous array of cells and is applied to the first input in a recurring manner; a second input to the comparator, the second input alternating between a reference voltage and an analog input voltage, wherein the comparator generates an output voltage based on a comparison of at least two of the voltage ramp, the reference voltage, and the analog input voltage; and a gated counter for measuring pulse-widths of a signal that is based at least in part on the output voltage of the comparator, wherein the pulse-widths of the signal carry information about a magnitude of the analog input voltage.