Artificial neural networks are known for their ability to solve problems for which we do not have viable solutions or algorithms, but their application is restraint by the absence of a direct massively parallel implementation. This is a major drawback, particularly during the often time-consuming training phase. Recently, Genobyte Inc. has built the CAM-brain machine (CBM). The CBM is specialised, FPGA-based, reconfigurable (evolvable) hardware, capable of evaluating cellular automata based neural network modules directly in silicon at very high speed. The main purpose of this research is to determine whether the particular paradigm, as implemented in the CBM (or with slight modifications) is able to successfully combine the advantages of neural networks with the speed of a parallel hardware implementation. Ultimately we hope to implement a relevant and convincing application in the context of bioinformatics.