Resumen
For the past 20 years, parallel computation has helped solve many significant problems in applied science and engineering, most notably those in high-performance applications not implementable on sequential computers . However, parallel computing is still not used in all areas of computer science nor has it found a significant role in mainstream computing. Even if parallelism meets its technical goals, achieving the status of a mature technology in the next years remains a serious challenge. A major factor limiting parallel computation in mainstream computer science is the lack of general purpose parallel computing models. Parallel computing is a class of systems including many different architectures-from single-instruction-multiple-data (SIMD) machines to distributed-memory, multiple-instruction-multiple-data (MIMD) computers and workstation clusters. Therefore, it is difficult to define a unifying architectural model for parallel computing, in the same way the von Neumann model is the unifying model for sequential computing.