A CNN Universal Chip in Cmos Technology

Abstract

This paper describes the design of a programmable cellular neural network (CNN) chip with added functionalities similar to those of the CNN universal machine. The prototype contains 1024 cells and has been designed in a 1·0 μm, n‐well CMOS technology. Careful selection of the topology and design parameters has resulted in a cell density of 31 cells mm−2 and around 7–8 bits accuracy in the weight values. Adaptive techniques have been employed to ensure accurate external control and system robustness against process parameter variations.

One-dimensional discrete-time CNN with multiplexed template-hardware

“…Moreover, an innovative feature of our approach is an hardware-multiplexing scheme optimizing the number of multipliers required to implement templates A and B. A similar hardware multiplexing approach has been previously reported in [14]. However, an element of novelty is here introduced.…”

“…However, an element of novelty is here introduced. In [14], the hardware multiplexing ultimately halves the processing speed. Here we exploit an intrinsic idle phase of the system, available anyway because a full delay must be realized by cascading two half-delay cells, avoiding processing speed penalties.…”

Low-Power Smart Imagers for Vision-Enabled Sensor Networks

“…and where the losses term, introduced in the full-signal-range (FSR) CNN model [Espejo et al, 1996], has the form:…”

2D operators on topographic and non‐topographic architectures—implementation, efficiency analysis, and architecture selection methodology

“…In the implementation area, the first 10 years yielded more than a dozen CNN chips made by only a few designers. Some of them followed the original CNN architecture [6], others made slight modifications, such as the full signal range model [10] [11], or discrete time CNN (DTCNN) [7], or skipped the dynamics, and made dense threshold logic in the blackand-white domain [8] only. All of these chips had cellular architecture, and implemented the programmable A and/or the B template matrices of the CNN Universal Machine [3] [26] In the second decade, this community slightly shifted the focus of chip implementation.…”