On parallel computation based on the domain partitioning, the efficient method for communication between partitioned subdomains was investigated. In the method, the communication table was generated by dividing a communication process into multi-stage, limiting each core assigned with each subdomain to communicate once at each stage, and determining the core with more neighboring subdomains in priority to another to communicate until no more cores were available at each stage. The parallel computation of fluid flow based on the finite volume method was performed and it was found that the parallel computation with the proposed communication table could successfully reduce computational time for communication compared with that with the conventional one with an increase in the number of cores used.
We present a new model for distributed shared memory systems, based on remote data accesses. Such features are offered by network interface cards that allow one-sided operations, remote direct memory access and OS bypass. This model leads to new interpretations of distributed algorithms allowing us to propose an innovative detection technique of race conditions only based on logical clocks. Indeed, the presence of (data) races in a parallel program makes it hard to reason about and is usually considered as a bug.