Winner-Takes-All Associative Memory: A Hamming Distance Vector Quantizer

  • Philippe O. Pouliquen
  • Andreas G. Andreou
  • Kim Strohbehn


We present a design methodology for mapping neuralyinspired algorithms for vector quantization, into VLSI hardware.We describe the building blocks used: memory cells, current conveyors,and translinear circuits. We use the basic building blocks todesign an associative processor for bit-pattern classification;a high-density memory based neuromorphic processor. Operatingin parallel, the single chip system determines the closest match,based on the Hamming distance, between an input bit pattern andmultiple stored bit templates; ties are broken arbitrarily. Energyefficient processing is achieved through a precision-on-demandarchitecture. Scalable storage and processing is achieved througha compact six transistor static RAM cell/ALU circuit. The singlechip system is programmable for template sets of up to 124 bitsper template and can store up to 116 templates (total storagecapacity of 14 Kbits). An additional 604 bits of auxiliary storageis used for pipelining and fault tolerance re-configuration capability.A fully functional 6.8 mm by 6.9 mmchip has been fabricated in a standard single–poly, double–metal2.0µmn–well CMOS process.

Analog VLSI vector quantization associative memory computational memory smart memory associative processor Winner-Takes-All precision-on-demand current-mode subthreshold circuits 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    T. Kohonen, Content-Addressable Memories, 2 edition. Springer Verlag, Berlin, 1987.Google Scholar
  2. 2.
    T. Kohonen, Self-Organization and Associative Memory, 2 edition. Springer Verlag: Berlin, 1988.Google Scholar
  3. 4.
    M. R. Emerling M. A. Sivilotti and C. A. Mead, VLSI architectures for implementation of neural networks. In (J. J. Denker, ed.), Neural Networks for Computing. AIP: Snowbird UT, 1986, pages 408–413Google Scholar
  4. 5.
    K. A. Boahen and A. G. Andreou, “Design of a bidirectional associative memory chip,” in Associative Neural Memories: Theory and Implementation (M. Hassoun, ed.), chapter 17. Oxford University Press: New York, NY, 1993.Google Scholar
  5. 6.
    P. O. Pouliquen, A. G. Andreou, K. Strohbehn, and R. E. Jenkins, “An associative memory integrated system for character recognition,” in Proc. 36th Midwest Symp. on Circuits and Systems, Detroit, MI, Aug. 1993, pp. 762–765.Google Scholar
  6. 7.
    Y. He, U. Cilingiroglu, and E. Sánchez-Sinencio, “A high density and low-power charge-based hamming network.” IEEE Trans. VLSI Systems 1(1), pp. 55–62, March 1993.Google Scholar
  7. 8.
    Y. Horio and S. Nakamura, “Analog memories for VLSI neurocomputing,” in Artificial Neural Networks: Paradigms, Applications, and Hardware Implementations (E. Sánchez-Sinencio and C. Lau, eds.), IEEE Press, Piscataway, NJ, 1992, pp. 344–366.Google Scholar
  8. 9.
    G. Cauwenberghs, C. F. Neugebauer, and A. Yariv, “Analysis and verification of an analog VLSI incremental outer-product learning systems.” IEEE Trans. Neural Networks 3(3), May 1992.Google Scholar
  9. 10.
    H. Yang, B. J. Sheu, and J.-C. Lee, “A nonvolatile analog neural memory using floating-gate MOS transistors.” Analog Integrated Circuits and Signal Processing 2(1), Feb. 1992.Google Scholar
  10. 11.
    K. A. Boahen, P. O. Pouliquen, A. G. Andreou, and R. E. Jenkins, “A heteroassociative memory using current-mode MOS analog VLSI circuits.” IEEE Trans. Circuits and Systems CAS-36(5), pp. 747–755, May 1989.Google Scholar
  11. 12.
    K. A. Boahen, A. G. Andreou, and P. O. Pouliquen, “Architectures for associative memories using current-mode analog MOS circuits,” in Advanced Research in VLSI: Proc. Dec. Caltech Conference on VLSI (C. L. Seitz, ed.). MIT Press: Cambridge, MA, 1989.Google Scholar
  12. 13.
    Eric Vittoz and J. Fellrath, “CMOS analog integrated circuits based on weak inversion operation.” IEEE J. Solid-State Circ. 12, pp. 224–231, 1977.Google Scholar
  13. 14.
    C. A. Mead, Analog VLSI and Neural Systems. Addison-Wesley: Reading, MA, 1989.Google Scholar
  14. 15.
    A. G. Andreou and K. A. Boahen, “Neural information processing II,” in Analog VLSI Signal and Information Processing (M. Ismael and T. Fiez, eds.). McGraw-Hill, 1994.Google Scholar
  15. 16.
    K. Seno, K. Knorpp, L.-L. Shu, N. Teshima, H. Kihari, H. Sato, F. Miyaji, M. Takeda, M. Sasaki, Y. Tomo, P. T. Chuang, and K. Kobayashi, “A 9-ns 16-mb CMOS SRAM with offset-compensated current sense amplifier.” IEEE J. Solid-State Cicuits 28(11), pp. 1119–1124, November 1993.Google Scholar
  16. 17.
    K. Sasaki, K. Ueda, K. Takasugi, H. Toyoshima, K. Ishibashi, T. Yamanaka, N. Hashimoto, and N. Ohki, “A 16-mb CMOS SRAM with a 2.3-μm 2 single bit line memory cell.” IEEE J. Solid-State Cicuits 28(11), pp. 1125–1130, November 1993.Google Scholar
  17. 18.
    R. F. Lyon and R. R. Schediwy, “CMOS static memory with a new four-transistor memory cell,” in Advanced Research in VLSI (P. Losleben, ed.). MIT Press: Cambridge, MA, 1987, pp. 110–132Google Scholar
  18. 19.
    G. A. Carpenter and S. Grossberg, “A massively parallel architecture for a self-organizing neural pattern recognition machine.” Computer Vision, Graphics, and Image Processing 37, pp. 54–115, 1987.Google Scholar
  19. 20.
    J. Lazzaro, S. Ryckebusch, M. A. Mahowald, and C. Mead, “Winner-take-all networks of o(n) complexity,” in Advances in Neural Information Processing Systems 1 (D. Touretzky, ed.). Morgan Kaufmann: San Manteo, CA, 1989, pp. 703–711Google Scholar
  20. 21.
    B. Gilbert, “Translinear circuits: A proposed classification.” Electronics Letters 11(1), pp. 14–16, 136, January 9, 1975.Google Scholar
  21. 22.
    A. G. Andreou and K. A. Boahen, “Translinear circuits in subthreshold mos.” J. Analog Integrated Circ. Sig. Proc. 9, pp. 141–166, 1996.Google Scholar
  22. 23.
    A. Pavasović, A. G. Andreou, and C. R. Westgate, “Characterization of subthreshold MOS mismatch in transistors for VLSI systems.” Journal of Analog Integrated Circuits and Signal Processing 6, pp. 75–85, July 1994.Google Scholar
  23. 24.
    N. Kumar, P. O. Pouliquen, and A. G. Andreou, “Device mismatch limitations on the performance of an associative memory system,” in Proc. 36th Midwest Symp. on Circuits and Systems (IEEE, ed.), 1993, pp. 570–573.Google Scholar
  24. 25.
    H. Miwa, N. Kumar, P. O. Pouliquen, and A. G. Andreou, “Memory enhancement techniques for mixed digital memory-analog computational engines,” in Proc. IEEE Int. Symp. on Circuits and Systems, volume 5, June 1994, pp. 45–48.Google Scholar

Copyright information

© Kluwer Academic Publishers 1997

Authors and Affiliations

  • Philippe O. Pouliquen
    • 1
  • Andreas G. Andreou
    • 2
  • Kim Strohbehn
    • 1
  1. 1.Electrical and Computer Engineering, Center for Language and Speech ProcessingJohns Hopkins UniversityBaltimore
  2. 2.Applied Physics LaboratoryJohns Hopkins UniversityLaurelUSA

Personalised recommendations