Dataflow Architectures: Flexible Platforms for Neural Network Simulation

Abstract

Dataflow architectures are general computation engines optimized for the execution of fme-grain parallel algorithms. Neural networks can be simulated on these systems with certain advantages. In this paper, we review dataflow architectures, examine neural network simulation performance on a new generation dataflow machine, compare that performance to other simulation alternatives, and discuss the benefits and drawbacks of the dataflow approach.

Cite

Text

Smotroff. "Dataflow Architectures: Flexible Platforms for Neural Network Simulation." Neural Information Processing Systems, 1989.

Markdown

[Smotroff. "Dataflow Architectures: Flexible Platforms for Neural Network Simulation." Neural Information Processing Systems, 1989.](https://mlanthology.org/neurips/1989/smotroff1989neurips-dataflow/)

BibTeX

@inproceedings{smotroff1989neurips-dataflow,
  title     = {{Dataflow Architectures: Flexible Platforms for Neural Network Simulation}},
  author    = {Smotroff, Ira},
  booktitle = {Neural Information Processing Systems},
  year      = {1989},
  pages     = {818-825},
  url       = {https://mlanthology.org/neurips/1989/smotroff1989neurips-dataflow/}
}