Abstract The stochastic network calculus is a recent framework for computing performance metrics such as backlog and delay in terms of probabilistic bounds. The first part of the talk briefly overviews the main concepts and formalisms of the network calculus. The calculus is then applied to the analysis of end-to-end delays of a flow in a worst-case network scenario in which (1) the flow has the lowest scheduling priority at the nodes, and (2) statistical independence is not assumed between flows at different nodes. In this scenario end-to-end delays scale as $\Theta(H\log H)$, where $H$ is the number of traversed nodes. This is different from the $\Theta(H)$ scaling predicted in other frameworks such as product-form queueing theory. The last part of the talk gives some insight into the accuracy of network calculus delay bounds, which are shown to yield utilization regions that saturate the network capacity at high data rates once accounting for statistical independence. Bio Florin Ciucu was educated at the Faculty of Mathematics, University of Bucharest (B.Sc. in Informatics, 1998), George Mason University (M.Sc. in Computer Science, 2001), and University of Virginia (Ph.D. in Computer Science, 2007). He is currently a Postdoctoral Fellow in the Electrical and Computer Engineering at the University of Toronto. His research interests are in the performance analysis of communication networks using stochastic models. Florin is a recipient of the ACM Sigmetrics 2005 Best Student Paper Award.