A parallel computing architecture for computing particle interactions.

摘要

A generalized approach to particle interaction can confer advantages over previously described method in terms of one or more of communications bandwidth and latency and memory access characteristics. These generalizations can involve one or more of at least spatial decomposition, import region rounding, and multiple zone communication scheduling. In a parallel implementation, the generalized approach can have the advantage that the communication bandwidth required by several instances of the approach come close or achieve an approximate lower bound for communication bandwidth. In a serial processing implementation, effective cache utilization can substantially reduce memory latency in processing.