Evaluating microprocessor multimedia extensions for the real-time simulation of RBF networks

In this article we determine the performances of general purpose microprocessors with multimedia extensions, i.e., including a SIMD micro-machine. We studied the Pentium and Pentium II processors with MMX technology to evaluate the capability of the multimedia extensions to simulate in real-time RBF networks using the Mahalanobis distance. For that purpose we implemented the Mahalanobis distance both in MMX assembly code and in C language. Our experiments showed that the Pentium processors with MMX technology are capable of simulating RBF neural networks in real-time with a speed up greater than 3 over usual code in C.