The Connex Array™ as a neural network accelerator
Răzvan Andonie, Mihaela Maliţa · Computational intelligence · 2007
We discuss the parallel implementation of neural networks on a Connex Array™circuit. By estimating the number of memory cycles, we approximate the real performance of the machine. We show how to implement a basic neural learning algorithm. Our preliminary study suggests that the Connex Array™is a good neural network accelerator, due to its high performance for vector operations. The execution time obtained is significant: 5 GCUPS, at 2.5 watts, on less than 1 cm2 area.