Optimization of complex system architectures can help accelerate the non-biased search for novel architec- tures in the early design phase. Four aspects needed to enable architecture optimization are discussed: formalization of the architecture design space, systematic exploration of the design space, conversion from architecture model to simulation model, and flexible simulation of architecture performance. Modeling the design space is driven by system requirements and simulation capabilities and should be based on functional decomposition. Systematic exploration can be done using enumeration, design of experiments, or optimization. Various approaches for converting architectures to simulation models are discussed. Finally, simulation environments should expose a flexible and modular interface to be used in architecture optimization.