Parallel processing using networked resources
R. Ahmed · 2002
We optimize the execution time of generic software modules on various machine architectures using networked computing hardware. A four layered platform implements this parallel processing algorithm. The set-up layer computes the feasibility of networked code execution by comparing the approximated execution times on the local machine and networked machines. The platform would either return with a slight overhead or would execute the code over the network in less time. The slight overhead is the time expended on the feasibility calculation. A case study for the execution of the Mandelbrot fractal code concludes that the algorithm reduces the execution time by at least 3 times for one specific case.