The IT world was built on scale-up, or vertically-scaled computing—where all resources required for a particular workload or, typically, multiple workloads, were housed on one large, centrally managed system.
The proliferation of x86 servers in the 1990s, combined with the emergence of large Web server farms, helped popularize a new model: scale-out or horizontally-scaled computing, where workloads are distributed across networks of typically small, inexpensive 2- and 4-processor servers.
Scale-out continues to greatly outgrow the market with the emergence of new workloads and the relative ease of migrating some scale-out applications.
This report examines the real future of scale-out computing.
It explores the infrastructure technology trends that are driving the move to scale-out and how these trends are likely to play out over the next five years.
It examines the evolution of computing workloads, focusing on which ones can and should—and which can't and shouldn't—go to scale-out.
It then assesses the relative strengths and weaknesses of scale-out versus scale-up, how the two distinct architectures are beginning to converge and what this will mean to customers and vendors.
Read this Report to Learn:The server trends—especially around blade servers—that will simultaneously drive and restrain the adoption of scale-out; The scale-up/scale-out workload continuum and which applications will migrate to scale-out, how they will do so and over what timeframes; How the emergence of a new model, scale-across, or diagonal scaling, will force a total rethinking of the traditionally binary scale-up/out decision; and Which systems vendors are best positioned to capitalize on these trends and the gotchas that can trip them up.