Research

ES2 researchers work closely with industry partners to define research areas relevant to the challenges facing industry. This could mean finding ways to make their own data centers run more efficiently, or serving as a research arm for members who may benefit from incorporating our research into the products and services that they offer. We regularly seek input from members in defining research projects and in mentoring students doing the research. Our members have helped us to define several key focus areas or research themes. These include:

  • Data center sustainability, availability and system energy optimization

This theme looks at the holistic provisioning of data center resources and fault tolerance through intelligent control and management system, including incorporating AI, system learning and neural networks.

DATA CENTER ENVIRONMENTAL CALCULATOR. This simple, open calculator predicts the Scope 3 carbon usage effectiveness (CUE) and water scarcity usage effectiveness (WSUE, defined as water scarcity footprint per IT load), based on facility power usage effectiveness (PUE), water usage effectiveness (WUE), and geographic location. The calculator is built using an Excel workbook with a Visual Basic extension. The calculator enables users to determine the influence of location, PUE, and WUE on a data center’s carbon footprint and water scarcity footprint.

  • Holistic IT load management and control

Research in this area has resulted in algorithms to predict workloads and balance those workloads across the optimum number of servers that should be deployed at any point in time. The algorithms continue to be refined, controlling the right amount of cooling corresponding to the servers that are running.

  • Air cooling systems

Various aspects of air-cooling are studied and modeled, including airflow management, containment, active tiles, close-coupled cooling systems, contamination and evaporative cooling. We holistically control thermal management to handle workload dynamics through verified predictive modeling.

  • Control and Optimization

We use sense data (temperature, airflow and pressure) in real time to train and calibrate empirical and physics-based intelligent models.

  • Warm water liquid cooling

This area looks at energy efficiency improvements through the use of compressor-less cooling, control systems and optimization of cooling systems and components such as cold plates.

  • Two-phase and immersion cooling

This research area is investigating the performance of pumped and non-pumped indirect two-phase cooling systems with emphasis on advancing the performance of evaporators for high power servers, as well as direct immersion cooling.

  • Waste energy harvesting 

This area explores best practices for harvesting waste heat from servers towards reducing equipment electrical loads or data center cooling loads.

  • Power distribution and storage

Research in this area studies the use of control systems and high efficiency power electronics related to the use of stored electrical energy for first level backup and load capping in data centers.

  • Embedded cooling

This research area is focused on establishing the limits of advanced methods for direct-contact of air and liquid cooling within the IT equipment and its components at the chip and module level.