Mixed Digital–Analog Associative Memory Enabling Fully-Parallel Nearest Euclidean Distance Search
スポンサーリンク
概要
- 論文の詳細を見る
In this paper, an associative memory architecture with mixed digital–analog search circuitry is proposed, which enables a fully-parallel nearest match-data search based on the Euclidean distance between input pattern and stored reference patterns. A test chip, implementing the proposed architecture, was designed in 0.35 μm complementary metal–oxide–semiconductor (CMOS) technology with two-poly and three-metal layers. The nearest-match unit consumes only 0.64 mm2 (12.5% of the total design area), while the whole chip area is 5.12 mm2. The layout-based simulated winner-search time, the time to determine the best-matching reference-data word for an input-data word among a database of 64 reference patterns (5-bit, 16 units), is lower than 160 ns. This corresponds to a performance requirement of 27 giga operations per second (GOPS)/mm2, if a general purpose 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 38 mW/mm2 at this high processing performance.
- Published by the Japan Society of Applied Physics through the Institute of Pure and Applied Physicsの論文
- 2007-04-30
著者
-
Mattausch Hans
Research Center For Nanodevices And Systems Hiroshima University
-
Tanaka Yuki
Research Center For Nanodevices And Systems Hiroshima University
-
Abedin Md.
Research Center For Nanodevices And Systems Hiroshima University
-
Ahmadi Ali
Research Center For Nanodevices And Systems Hiroshima University
-
Koide Tetsushi
Research Center For Nanodevices And Systems Hiroshima University
-
Abedin Md.
Research Center for Nanodevices and Systems, Hiroshima University, 1-4-2 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8527, Japan
-
Mattausch Hans
Research Center for Nanodevices and Systems, Hiroshima University, 1-4-2 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8527, Japan
-
Koide Tetsushi
Research Center for Nanodevices and Systems, Hiroshima University, 1-4-2 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8527, Japan
-
Tanaka Yuki
Research Center for Nanodevices and Systems, Hiroshima University, 1-4-2 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8527, Japan
関連論文
- 4-Port Unified Data/Instruction Cache Design with Distributed Crossbar and Interleaved Cache-Line Words(Integrated Electronics)
- Acceleration of DCT Processing with Massive-Parallel Memory-Embedded SIMD Matrix Processor(Image Processing and Video Processing)
- Realization of K-Nearest-Matches Search Capability in Fully-Parallel Associative Memories(VLSI Design Technology and CAD)
- Mixed digital-analog associative memory enabling fully-parallel nearest euclidean distance search (Special issue: Solid state devices and materials)
- A Performance-Driven Floorplanning Method with Interconnect Performance Estimation(Special Section on VLSI Design and CAD Algorithms)
- Circuit-Simulation Model of C_ Changes in Small-Size MOSFETs Due to High Channel-Field Gradients(the IEEE International Conference on SISPAD '02)
- A Compact Model of the Pinch-off Region of 100nm MOSFETs Based on the Surface-Potential(Semiconductor Materials and Devices)
- 1/f-Noise Characteristics in 100 nm-MOSFETs and Its Modeling for Circuit Simulation(Semiconductor Materials and Devices)
- Quantum Effect in Sub-0.1μm MOSFET with Pocket Technologies and Its Relevance for the On-Current Condition
- Circuit Simulation Models for Coming MOSFET Generations(Special Section of Selected Papers from the 14th Workshop on Circuits and Systems in Karuizawa)
- Integration Architecture of Content Addressable Memory and Massive-Parallel Memory-Embedded SIMD Matrix for Versatile Multimedia Processor
- Scalable FPGA/ASIC Implementation Architecture for Parallel Table-Lookup-Coding Using Multi-Ported Content Addressable Memory(Image Processing and Video Processing)
- Real-Time Huffman Encoder with Pipelined CAM-Based Data Path and Code-Word-Table Optimizer(Image Processing and Video Processing)
- A CAM-Based Signature-Matching Co-processor with Application-Driven Power-Reduction Features(Integrated Electronics)
- Digital Low-Power Real-Time Video Segmentation by Region Growing
- 100 nm-MOSFET Model for Circuit Simulation : Challenges and Solutions(Devices and Circuits for Next Generation Multi-Media Communication Systems)
- Analysis of Within-Die Complementary Metal--Oxide--Semiconductor Process Variation with Reconfigurable Ring Oscillator Arrays Using HiSIM
- MOSFET Harmonic Distortion up to the Cutoff Frequency : Measurement and Theoretical Analysis
- 100 nm-MOSFET Model for Circuit Simulation : Challenges and Solutions
- Fast and Compact Central Arbiter for High Access-Bit-Rate Multi-Port Caches
- Bank-Type Associative Memory for High-Speed Nearest Manhattan Distance Search in Large Reference-Pattern Space
- Efficient Video-Picture Segmentation Algorithm for Cell-Network-Based Digital CMOS Implementation(Image Processing, Image Pattern Recognition)
- High Speed Frequency-Mapping-Based Associative Memory Using Compact Multi-Bit Encoders and a Path-Selecting Scheme (Special Issue : Solid State Devices and Materials (2))
- Low-Power Silicon-Area-Efficient Image Segmentation Based on a Pixel-Block Scanning Architecture
- Software-Based Parallel Cryptographic Solution with Massive-Parallel Memory-Embedded SIMD Matrix Architecture for Data-Storage Systems
- A K-Means-Based Multi-Prototype High-Speed Learning System with FPGA-Implemented Coprocessor for 1-NN Searching
- Experimental Analysis of Within-Die Process Variation in 65 and 180 nm Complementary Metal--Oxide--Semiconductor Technology Including Its Distance Dependences
- Mixed Digital–Analog Associative Memory Enabling Fully-Parallel Nearest Euclidean Distance Search
- Automatic Pattern-Learning Architecture Based on Associative Memory and Short/Long Term Storage Concept