Associative memory with fully parallel nearest-Manhattan-distance search for low-power real-time single-chip applications

Authors:
Yuji Yano;Tetsushi Koide;Hans Jürgen Mattausch
Affiliations:
Hiroshima University, Kagamiyama, Higashi-Hiroshima, Japan;Hiroshima University, Kagamiyama, Higashi-Hiroshima, Japan;Hiroshima University, Kagamiyama, Higashi-Hiroshima, Japan
Venue:
Proceedings of the 2004 Asia and South Pacific Design Automation Conference
Year:
2004

Citing 2
Cited 2

Vector quantization and signal compression

Vector quantization and signal compression
The Pattern Recognition Basis of Artificial Intelligence

The Pattern Recognition Basis of Artificial Intelligence

A hardware design of a massive-parallel, modular NN-based vector quantizer for real-time video coding

Microprocessors & Microsystems
K-means clustering algorithm for multimedia applications with flexible HW/SW co-design

Journal of Systems Architecture: the EUROMICRO Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

A fully-parallel minimum Manhattan-distance search associative memory has been designed in 0.35μm CMOS with 3-metal layers. The nearest-match unit consumes only 1.02mm2, while the chip area is 7.49mm2. The measured winner-search time of this chip, the time to determine the best-matching reference-data word for an input-data word among a database of 128 reference words (5-bit, 16 units), is 2, if a 32-bit computer with the same chip area would have to run the same workload. Furthermore the power dissipation of the designed test chip is only about 26.7mW/mm2.