Beschreibung
Although computer networks are inherently parallel systems, the parallel execution of network simulations on interconnected processors frequently yields only limited benefits. In this thesis, methods are proposed to estimate and understand the parallelization potential of network simulations. Further, mechanisms and architectures for exploiting the massively parallel processing resources of modern graphics cards to accelerate network simulations are proposed and evaluated.
Herstellerkennzeichnung:
BoD - Books on Demand
In de Tarpen 42
22848 Norderstedt
DE
E-Mail: info@bod.de
