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

Massively parallel architectures for large scale neural network simulations.

Y Fujimoto1, N Fukuda, T Akabane

  • 1Sharp Corp., Nara.

IEEE Transactions on Neural Networks
|January 1, 1992
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

Direct observation of the superallowed α-decay of <sup>104</sup>Te.

Nature·2026
Same author

Dipole Strength Distribution of ^{8}He and Decay Characteristics.

Physical review letters·2026
Same author

β-Decay Half-Lives of Neutron-Rich Sulfur to Potassium: Evolution of the N=32 and 34 Subshell Closures below Calcium.

Physical review letters·2025
Same author

Precision Mass Measurements around ^{84}Mo Rule Out ZrNb Cycle Formation in the Rapid Proton-Capture Process at Type I X-Ray Bursts.

Physical review letters·2025
Same author

Seniority Structure in Neutron-Rich Nucleus ^{128}Ag: Evidence for Robustness of N=82 Shell Closure in Silver Isotopes.

Physical review letters·2025
Same author

Isospin Symmetry Breaking in the ^{71}Kr and ^{71}Br Mirror System.

Physical review letters·2025
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

New toroidal lattice architecture (TLA) and planar lattice architecture (PLA) offer scalable neurocomputer designs for large-scale neural network simulations. Performance scales with processor count, addressing key challenges in parallel processing.

Area of Science:

  • Computer Science, Artificial Intelligence, Neuroscience

Background:

  • Large-scale neural network simulations require efficient parallel processing architectures.
  • Existing architectures face challenges with connectivity, data transmission bottlenecks, and load balancing.
  • Scalability and performance degradation are critical issues in neurocomputing.

Purpose of the Study:

  • To propose novel massively parallel neurocomputer architectures: Toroidal Lattice Architecture (TLA) and Planar Lattice Architecture (PLA).
  • To address connectivity, data transmission, and load balancing problems in large-scale simulations.
  • To demonstrate performance scalability proportional to the number of node processors.

Main Methods:

  • Development of TLA and PLA with efficient 2D processor connections for Wafer-Scale Integration (WSI).

Related Experiment Videos

  • Definition of a general neuron model.
  • Implementation of TLA using transputers, including Hopfield networks and multilayer perceptrons.
  • Application to the traveling salesman problem and identity mapping.

    • Main Results:

    • Architectures demonstrate performance nearly proportional to the number of node processors.
    • Successful implementation and application of neural network models on the proposed architectures.
    • Validation of solutions for connectivity, data transmission, and load balancing challenges.

    Conclusions:

    • TLA and PLA provide efficient and scalable solutions for large-scale neural network simulations.
    • The proposed architectures effectively overcome performance degradation issues.
    • The findings confirm the potential for significant advancements in neurocomputing hardware.