“…For architectural exploration, the HLM is typically augmented with architecture information, i.e., number and type of processing and communication resources including buses and memories. By specifying the HLM mapping to this architecture, and by using estimates for execution times, so called Virtual Architecture Models can be generated, permitting a combined functional and timed simulation, and hence, an early performance estimation [2], [3]. After selecting the platform (i.e., an architecture instance) for implementation, the HLM can be reused in various scenarios: In hardware design, the HLM serves as golden model; in software design, parts of the executable HLM are integrated into a virtual prototype, allowing software development to start without the need for a hardware prototype.…”