Enterprise distributed real-time and embedded (DRE) systems are increasingly using high-performance computing architectures, such as dual-core architectures, multi-core architectures, and parallel computing architectures, to achieve optimal performance. Performing system integration tests on such architectures in realistic operating environments during early phases of the software lifecycle, i.e., before complete system integration time, is becoming more critical. This helps distributed system developers and testers evaluate and locate potential performance bottlenecks before they become too costly to locate and rectify. Traditional approaches either (1) rely heavility on simulation techiques or (2) are too low-level and fall outside the domain knowledge distributed system developers and testers. Consequently, it is hard for distributed system developers and testers to produce realistic operating conditions for early integration testing of such systems.

This papers provides two contributions to facilitating early system integration testing of enterprise DRE systems. First, it provides a generalized technique for emulating computation intensive workload irrespective of the target architecture. Secondly, this paper illustrates how the emulation technique is used to evaluating different high-performance computing architectures in early phases of the software lifecycle. The technique presented in this paper is empirically and quantitatively evaluated in the context of a representative enterprise DRE system from the domain of shipboard computing environments.