Achieving scalability in computer systems without sacrificing usability, requires a synergistic combination of system architecture, fundamental technologies, and implementation. This paper discusses how the Silicon Graphics Origin system utilizes these elements to create a truly scalable microprocessor and makes predictions of how these elements will evolve to provide performance growth into the next century. The paper begins by reviewing current multiprocessor alternatives, and producing the notion of a scalable SMP. The second part of the paper focuses on a particular instance of a scalable SMP, the Silicon Graphics Origin multiprocessor and its S/sup 2/MP memory architecture. We give an overview of the Origin system architecture and then discuss some of the core technologies and key implementation components of the system. In the final section we examine how technology trends will impact system architecture and what key technologies and implementation strategies are implied by those trends. We go on to predict that clusters of scalable shared-memory multiprocessors will become the dominant, multiprocessor architecture.