Trusted computing (e.g. TCPA and Microsoft’s NextGeneration Secure Computing Base) has been one of the most talked about and least understood technologies in the computing community over the past year. The capabilities trusted computing provides have the potential to radically improve the security and robustness of distributed systems. Unfortunately, the debate over its application to digital rights management has caused its significant other applications to be largely overlooked. In this paper we present a broader vision for trusted computing. We give an intuitive model for understanding the capabilities and limitations of the mechanisms provided by trusted computing. We describe a flexible OS architecture to support trusted computing. We present a range of practical applications that illustrate how trusted computing can be used to improve security and robustness in distributed systems.