Exploiting SIMD computers for general purpose computation

This paper proposes a strategy for exploiting massively parallel SIMD computers for general purpose computation. The approach places compiled programs into the local memory space of each distinct processing element (PE). Within each PE, a local program counter is initialized and the instructions are interpreted in parallel across all of the PEs by control signals emanating from the central control unit. Initial experiments with randomly generated programs show that speedup of approximately 700 is attainable on a SIMD processor with 8 K processing elements. Furthermore, additional experiments have shown that the speedup increases linearly with the number of processing elements