Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A massively parallel architecture for self-organizing feature maps.

M Porrmann1, U Witkowski, U Ruckert

  • 1Heinz Nixdorf Inst., Paderborn Univ., Germany.

IEEE Transactions on Neural Networks
|February 5, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Coherent optical communication: towards realtime systems at 40 Gbit/s and beyond.

Optics express·2008
Same author

[Rare indication for a simple hip prosthesis with an extra-large femoral head following multiple failed exchange operations].

Der Chirurg; Zeitschrift fur alle Gebiete der operativen Medizen·1999
Same author

[Wise women and pious sisters. Female physicians and nurses in the late middle ages].

Pflege aktuell·1995
Same author

[Elizabeth of Thuringia--the duchess in a rag dress].

Pflege aktuell·1995
Same author

[Hildegarde von Bingen, trumpet of the Lord].

Pflege aktuell·1995
Same author

[Hygieia, the goddess of health].

Pflege aktuell·1995
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study introduces a hardware accelerator for self-organizing feature maps (SOFMs). The massively parallel architecture efficiently implements SOFMs for embedded systems and large-scale simulations, accelerating neural network applications.

Area of Science:

  • Computer Engineering
  • Artificial Intelligence
  • Hardware Acceleration

Background:

  • Self-organizing feature maps (SOFMs) are crucial for unsupervised learning and data visualization.
  • Implementing SOFMs in hardware offers potential for significant performance gains.
  • Existing solutions may face limitations in scalability or resource efficiency for embedded applications.

Purpose of the Study:

  • To present a novel hardware accelerator for self-organizing feature maps.
  • To design a massively parallel architecture for efficient SOFM implementation.
  • To accelerate both SOFM operations and associated data preprocessing tasks.

Main Methods:

  • Development of a massively parallel hardware architecture.
  • Implementation of an Application-Specific Integrated Circuit (ASIC) for verification.

Related Experiment Videos

  • Integration of the ASIC into a heterogeneous multiprocessor system for neural applications.
  • Performance analysis across various simulation parameters.

    • Main Results:

    • The architecture enables resource-efficient implementation of small to medium-sized SOFMs for embedded systems.
    • Scalable simulation of large SOFMs is achievable using parallelized integrated circuits.
    • Hardware acceleration extends to data pre- and postprocessing, enhancing overall system efficiency.

    Conclusions:

    • The developed hardware accelerator offers a flexible and efficient solution for SOFM implementation.
    • The ASIC verification confirms the architectural concept's viability in real-world neural applications.
    • The proposed architecture demonstrates potential for high performance, even with future microelectronic technologies.