Search for: All records

In recent years, there have been phenomenal increases in Artificial Intelligence and Machine Learning that require data collection, mining and using data sets to teach computers certain things to learn, analyze image and speech recognition. Machine Learning tasks require a lot of computing power to carry out numerous calculations. Therefore, most servers are powered by Graphics Processing Units (GPUs) instead of traditional CPUs. GPUs provide more computational throughput per dollar spent than traditional CPUs. Open Compute Servers forum has introduced the state-of-the-art machine learning servers “Big Sur” recently. Big Sur unit consists of 4OU (OpenU) chassis housing eight NVidia Tesla M40 GPUs and two CPUs along with SSD storage and hot-swappable fans at the rear. Management of the airflow is a critical requirement in the implementation of air cooling for rack mount servers to ensure that all components, especially critical devices such as CPUs and GPUs, receive adequate flow as per requirement. In addition, component locations within the chassis play a vital role in the passage of airflow and affect the overall system resistance. In this paper, sizeable improvement in chassis ducting is targeted to counteract effects of air diffusion at the rear of air flow duct in “Big Sur” Open Compute machine learning server wherein GPUs are located directly downstream from CPUs. A CFD simulation of the detailed server model is performed with the objective of understanding the effect of air flow bypass on GPU die temperatures and fan power consumption. The cumulative effect was studied by simulations to see improvements in fan power consumption by the server. The reduction in acoustics noise levels caused by server fans is also discussed. 

The objective of this work is to introduce and evaluate a new end-of-aisle cooling design which consists of three cooling configurations. The key objectives of close-coupled cooling are to enable controlled cooling of information technology (IT) equipment, flexible and modular design, and the containment of hot air exhaust from the cold air. The thermal performance of the proposed solution is evaluated using computational fluid dynamics modeling. A computational model of a small size data center room has been developed. The room is modeled to be a hot aisle containment setup, i.e., the hot air exhaust exiting for each row is contained and directed within a specific volume. The cold aisle is separated from the hot aisle by means of banks of heat exchangers (HXs) placed on either side of the containment aisle. Based on the placement of rack fans, the design is divided into three sub-designs—Case 1: passive HXs with rack fan walls; Case 2: active HXs (coupled with fans) with rack fan walls; Case 3: active HXs (coupled with fans) with no rack fans. The cooling performance is calculated based on the thermal and flow parameters obtained for all three configurations. The computational data obtained has shown that the Case 1 is used only for lower system resistance IT. However, Cases 2 and 3 can handle denser IT systems. Case 3 is the design that can consume lower fan energy and handle denser IT systems. The article also discusses the cooling behavior of each type of design under cooling failure conditions with Case 2 showing better cooling redundancy compared with other two cases. 

One of the key areas in which electronic cooling research has been focusing on, is addressing the issue of non-uniform at package level. This challenge has incited the use of numerous temperature sensing mechanisms for dynamic cooling of electronic components. What dynamic liquid cooling effectively does is, use feedback from sensors as inputs for the pumps, supplying more amounts of fluid to parts of the electronics that is warmer while supplying minimal fluid to the parts of the electronics that are relatively cooler. A novel approach to address uneven heating in a liquid cooled system is the use of a temperature sensing flow control device that can control flow rate based on temperature. The necessity of numerous temperature and pressure sensors, a suitable control system and the maintenance and reliability issues that they present, can be significantly minimized with the use of a self sensing and controlling flow control device. This paper looks at the flow analysis of a self-regulating flow control device (FCD) designed for electronic module for data center application. An axially rotating butterfly valve is used to regulate the flow rate of FCD. Linearization of the flow with respect to damper angle is studied by modifying the dimensional ratios of the rectangular cross section of the FCD. Pressure drop, and flow rate characterization is done for the FCD.