A Neurocomputer Board Based on the ANNA Neural Network Chip

Abstract

A board is described that contains the ANN A neural-network chip, and a DSP32C digital signal processor. The ANNA (Analog Neural Network Arithmetic unit) chip performs mixed analog/digital processing. The combination of ANNA with the DSP allows high-speed, end-to-end ex(cid:173) ecution of numerous signal-processing applications, including the prepro(cid:173) cessing, the neural-net calculations, and the postprocessing steps. The ANNA board evaluates neural networks 10 to 100 times faster than the DSP alone. The board is suitable for implementing large (million con(cid:173) nections) networks with sparse weight matrices. Three applications have been implemented on the board: a convolver network for slant detection of text blocks, a handwritten digit recognizer, and a neural network for recognition-based segmentation.

Cite

Text

Säckinger et al. "A Neurocomputer Board Based on the ANNA Neural Network Chip." Neural Information Processing Systems, 1991.

Markdown

[Säckinger et al. "A Neurocomputer Board Based on the ANNA Neural Network Chip." Neural Information Processing Systems, 1991.](https://mlanthology.org/neurips/1991/sackinger1991neurips-neurocomputer/)

BibTeX

@inproceedings{sackinger1991neurips-neurocomputer,
  title     = {{A Neurocomputer Board Based on the ANNA Neural Network Chip}},
  author    = {Säckinger, Eduard and Boser, Bernhard E. and Jackel, Lawrence D.},
  booktitle = {Neural Information Processing Systems},
  year      = {1991},
  pages     = {773-780},
  url       = {https://mlanthology.org/neurips/1991/sackinger1991neurips-neurocomputer/}
}