Search

Complete immersion of servers in synthetic dielectric fluids is rapidly becoming a popular technique to minimize the energy consumed by data centers for cooling purposes. In general, immersion cooling offers noteworthy advantages over conventional air-cooling methods as synthetic dielectric fluids have high heat dissipation capacities which are roughly about 1200 times greater than air. Other advantages of dielectric fluid immersion cooling include even thermal profile on chips, reduction in noise and addressing reliability and operational enhancements like whisker formation and electrochemical migration. Nevertheless, lack of data published and availability of long-term reliability data on immersion cooling is insufficient which makes most of data centers operators reluctant to implement this technique. The first part of this paper will compare thermal performance of single-phase oil immersion cooled HP ProLiant DL160 G6 server against air cooled server using computational fluid dynamics on 6SigmaET®. Focus of the study are major components of the server like Central Processing Unit (CPU), Dual in Line Memory Module (DIMM), Input/output Hub (IOH) chip and Input/output controller Hub (ICH). The second part of this paper focuses on thermal performance optimization of oil immersion cooled servers by varying inlet oil temperature, flow rate and using different fluid.

The energy consumption in a typical data center is rising day by day to meet the increasing demand of high compute servers resulting in high energy costs. In typical data centers, almost half of the IT energy is dedicated to cooling of the IT server racks. Hybrid cooling technology as an approach to minimize the power consumption uses of water or water based fluid to cool a high heat generating component whereas the rest of the components is cooled by air using internal fans. In this paper, our objective is to optimize the flow rate of air cooling loop of such hybrid cooled server to make data center cooling more cost effective and energy efficient. The volume of air supplied is controlled by varying the air flow rate through the internal fans. Also, the number of fans is reduced from five to three to minimize the power consumption. This finding helped us realize the 40% decrease in Fan power for each server. The pressure difference between the cold and hot aisle is created to avoid the backflow of cold air. Parameters like CPU and memory utilization were varied and temperature of different components such as processors, Dual-In-Line-Memory Module (DIMMs) & Platform Controller Hub (PCH) was monitored. When the server is running with 3 fans (approx. 10 CFM) maximum temperatures are found below the critical temperature. Further studies are also carried out through CFD analysis to observe changes in a model in 6SigmaET to increase the cooling efficiency. The study gives an idea of power savings by fan consolidation for energy efficient hybrid servers.

In response to the growing demand for power and energy across US and world-wide, development of energy efficient solutions has become very important. Considering data center applications, cooling power consumption constitutes a significant part of the overall energy usage of the system. In the process of optimizing the energy consumed per performance unit, liquid cooling has become one of the key solutions. Liquid cooling addresses the critical issues related to typical air cooling in servers because of its better heat transfer characteristics. Water-cooling at the device level can be an efficient solution since water has higher thermal capacitance when compared to traditional heat carrying medium i.e., air. Additionally, moving the cooling source closure to the heat load helps in reducing hot spots at the room level, thereby, improving the effectiveness of the cooling solution. On the other hand, the feasibility to operate liquid cooling at wider temperature range helps in partially or eliminating the room air conditioning units and chiller compressors and reduces the cooling power consumptions substantially. The primary objective of the study is to evaluate air vs. liquid cooled enterprise-class server for the effects of high ambient inlet temperatures. The air-cooled server has five hot swappable fans and two CPU heat sinks to dissipate the total heat in the server. The water-cooled counter-part consists of two-cold plates with integrated pumps to cool the CPUs while the dual in-line memory modules transfer the heat using a heat transfer tape attached to the manifold that is carrying the coolant. The existing practice in the industry is to maximize the use of economizer usage by reducing/eliminating the usage of chiller while taking advantage of outside ambient conditions to cool the data centers. However, higher ambient temperatures would lead to higher operating temperatures of the processors. The operating temperature of the CPU has direct influence on the static power which is known to reduce the performance of the processor. In this study, theair as well as liquid cooled servers will be tested at inlet temperatures ranging from 25°C-45°C (ASHRAE T.C 9.9 A4 and W4 classes) and the corresponding thermal and flow parameters will be compared for both cases. The current work primarily focuses on evaluating the critical inlet temperatures (for air cooled vs. liquid cooled server) at which CPU die temperatures has an influence of the static power. Collectively, a case will be presented as to why l iquid cooling would provide a viable option to operate the data centers at wider inlet temperatures with reduced risk of leakage current and sustain the performance and reliability of the IT equipment.