We present a technique that facilitates early prediction of software reliability and performance at the architecture design stage. Software architecture design is a crucial stage in the software development process, especially in developing large scale software. Early prediction of the reliability and performance of the software can be used as a basis for making design decisions. We have studied several common architecture styles and the impact of different configurations for the same style on reliability and performance measurements. The results from our studies show that altering architecture configuration to attain higher reliability and/or better performance is feasible. Furthermore, we propose a simple model that can be used in estimating reliability and performance measurement of software architectures.