An advanced energy efficient rack server design

Abstract

Data centers concern not just energy usage, but also other important overall considerations such as the actual computational work, the energy efficiency and the performance of servers. The server, as one of the key ingredients of a data center, plays an increasingly crucial role in contributing to the overall energy use, especially in cases where the efficiency of the infrastructure has been optimized. Baidu has been exploring the sweet zone between power and performance in efficient rack server design and deployment for their self-built data center energy efficiency optimization. Two types of research had deployed cases and continued to move forward: “Type 1, Power optimization”, power reduction without performance degradation; “Type 2, Energy proportional scaling”, server energy use scaled proportionally to workload, where the goal is to bring server energy efficiency closer to the optimal performance and power based on the specific workload characteristics. This paper addresses Baidu energy efficient rack server detail design, lab experiments result and deployment data surrounding both “Type 1, Power optimization” and “Type 2, Energy proportional scaling” research directions. One design case for each research direction is introduced in detail and energy efficiency contribution to Total Cost Ownership (TCO) saving as well as design case Best Known Method (BKM) are summarized. Future deployment plan of Baidu energy efficient server design is presented at the end. This paper also demonstrates how the efficiency of servers can be optimized for specific workloads as well as tailored to server configuration for optimum performance and efficiency.