Real world computing program - massively parallel systems and neural systems
Tatsumi Furuya · 2005
Real-world computing requires a computation framework that can process varied information and integrate it flexibly. Several parallel computation frameworks have been proposed, including concurrent object orientation, data parallel processing, neural network, and probability based information processing. These computation frameworks are suited to a massively parallel architecture, and neural systems involve enormous computations to solve practical problems within a reasonable time. In the first stage, massively parallel systems and neural systems are studied in parallel with some interaction between them. In the final stage, massively parallel systems and neural systems may be integrated into a new computer architecture.