Enterprise data centers consume an alarmingly-high fraction of the energy produced in the United States. The US Environmental Protection Agency estimates that data center energy consumption will reach over 100 billion kWh by 2011, 2.5% of US power generation, resulting in an estimated annual electricity cost of $7.4 billion. As much as 40% of this energy is wasted because of two key inefficiencies: (1) the substantial energy used by idle equipment that is powered on, but not performing useful work, and (2) inefficiency in data-center cooling infrastructure arising from a poor match between where heat is generated and where cool air is supplied. This project proposes research on a data-center-wide management system that controls IT equipment, power, and cooling infrastructure in real time to save energy in two ways. First, it actively consolidates computing tasks onto fewer systems, allowing idle systems to be powered down. Second, it moves computing tasks to systems that can be cooled most efficiently. The project addresses the key challenge of modeling temperature in large-scale data centers and designing an autonomous system that makes correct decisions on where to place tasks.
Computer architects and circuit designers have made enormous strides in managing the energy efficiency and peak power demands of processors and other silicon systems. Sophisticated power management features and modes are now myriad across system components. And yet, despite these advances, typical data centers today suffer embarrassing energy inefficiencies: it is not unusual for less than 20% of a data center's multi-megawatt total power draw to flow to computer systems actively performing useful work. Managing power and energy is challenging because individual systems and entire facilities are conservatively provisioned for rare utilization peaks, which leads to energy waste in underutilized systems and over-provisioning of physical infrastructure. Power management is particularly challenging for interactive Internet services like social networking, web search, ad serving, and machine translation that perform significant computing over massive data sets for each user request but require responsiveness in sub-second time scales. Data center inefficiencies lead to worldwide energy waste measured in billions of dollars and tens of millions of metric tons of CO2. he total carbon footprint of the world's data centers is estimated to roughly match the CO2 emissions of the entire Czech Republic. In the US, the Environmental Protection Agency estimates that data center energy consumption exceeds 100 billion kWh per year, over 2% of domestic power generation (more than the nation's color televisions), resulting in an estimated annual electricity cost exceeding $7 billion. Prof. Thomas Wenisch at the University of Michigan and his NSF-funded research team has developed a series of technologies designed to conserve power data centers. A great deal of energy is wasted in idle systems—in typical deployments, server utilization is below 30%, but idle servers still consume 60% of their peak power draw. Though frequent, typical idle periods last seconds or less, confounding simple energy-conservation approaches. The team has developed PowerNap, a scheme wherein servers transition rapidly between a high-performance active state and a near-zero-power idle state in response to instantaneous load, eliminating server idle-power waste. The Michigan research demonstrates that PowerNap provides greater energy efficiency and lower response time penalties than current solutions based on voltage and frequency scaling. Under the low average utilizations observed in typical data centers, PowerNap can reduce data center power costs by up to 75%.
