Abstract

In this talk I will explain the rational for the fundamental design decisions in the Cell Broadband Engine (CBE) and explain what these choices mean for programming Cell and for Cell applications. Because CBE provides a step-function in performance, it has allowed a number of applications to make the transition from off-line to real-time processing. I will highlight a number of such applications, focusing on areas other than medical imaging. Finally I will talk about how Cell is likely to develop in the future, and discuss application areas beyond those where Cell is used today where Cell can be expected to have an impact.