Application of the ANNA neural network chip to high-speed character recognition
Eduard Säckinger, Bernhard E. Boser, Jane M. Bromley, Yann LeCun, L. D. Jackel · IEEE Transactions on Neural Networks · 1992
A neural network with 136000 connections for recognition of handwritten digits has been implemented using a mixed analog/digital neural network chip. The neural network chip is capable of processing 1000 characters/s. The recognition system has essentially the same rate (5%) as a simulation of the network with 32-b floating-point precision.