Performance Prediction Based on Knowledge of Prior Product Versions

Performance estimation is traditionally carried out when measurement from a product can be obtained. In many cases there is, however, a need to start to make predictions earlier in a development project, when for example different architectures are compared. In this paper, two methods for subjective predictions of performance are investigated. With one of the methods experts estimate the relative resource usage of software tasks without using any knowledge of earlier versions of the product, and with the other method experts use their experience and knowledge of earlier versions of the system.With both methods there are rather large differences between different individual predictions, but the median of the prediction error indicates that the second method is worth further investigations.