A low-cost neuroprocessor board for emulating the SOFM neural model

Abstract

In this paper, the design and implementation of a low-cost PC-board for emulating the self-organising map neural model is shown. This neuroprocessor, based on Xilinx's FPGAs, is built and tested, and its performance is analysed and compared to computer simulations. The resulting neuroboard is just a first step towards the development of a general purpose neuroemulator based on FPGAs, in which the reconfiguration properties of these programmable logic devices will be broadly exploited.

General purpose neuroemulator architecture: design and VHDL simulation

“…In a previous work we showed a neuroprocessor board for emulating the SOFM neural model that we built by using FPGA chips [2]. However, a hardware architecture for general purpose neural network emulation is presented in the present paper.…”

“…6), in a XC4006-3 Xilinx's FPGA (this AU operates at 15,5 MHz), and we are now developing the local control unit. FPGA implementation allows modifying the structure of the different units in order to fitting the requirements of every neural model [2], feature that we will use for developing a reconfigurable general purpose neural processor. …”

“…With the term 'general purpose' we indicate the possibility of emulating a broad set of ANN models. Our ultimate goal is developing a general purpose reconfigurable neuroemulator, based on the principles tested in the prototype built for the SOFM [2]. This paper is organised as follows.…”