How to build and run edge data centres sustainably?

Download Listen

Sustainability is becoming increasingly crucial for data centres. Data centers currently consume around 1.5% of the world’s electricity, a figure expected to rise significantly (up to 10% in some countries) with the expansion of cloud and edge computing. This places strain on electricity grids and national climate objectives. As such, governments worldwide are introducing restrictions, limiting new data connections to the power grid. To continue operating, data centre operators must implement sustainable solutions in their facilities.

Edge sustainability: An overlooked aspect

While traditional data center sustainability receives substantial attention, edge computing sustainability often gets overlooked. STL Partners projects that the global market for edge computing is projected to grow from $9 billion in 2020 to over $450 billion by 2030. Ensuring this expansion is done in a sustainable manner is crucial. However, edge environments present unique challenges that require tailored strategies beyond those used in traditional data centers.

Edge computing environments differ significantly from traditional data centers. They often lack advanced heating, ventilation and air conditioning (HVAC) systems, uninterruptible power systems (UPS), and continuous support staff. As edge sites are often more remote and distributed, they require unique sustainability strategies.

Specialised cooling applied to COTS hardware: a solution to edge sustainability

Adopting traditional data centre sustainability strategies, such as renewable energy use, green code and waste energy reuse, for edge environments is the first step data centre operators should take. However, to advance their sustainability further, edge operators should focus on adopting more efficient cooling systems, as these currently can account for up to 40% of the energy used by a typical data centre.

Applying specialised cooling to COTS hardware is an advanced technique for managing the heat generated by computing equipment. Unlike traditional air-cooling methods, which rely on fans and air conditioning systems, specialised cooling uses liquids to absorb and dissipate heat more efficiently. This method is particularly effective in dense computing environments where traditional cooling solutions may fall short.

Types of specialised cooling solutions

Specialised cooling applied to COTS hardware can be broadly broken down into four subcategories:

• Liquid precision cooling: In this system, a coolant is delivered directly to the hottest components of the server before cascading it into the chassis to cool other components. The heated liquid is then transferred to a heat exchanger where the heat is dissipated into the air. This method eliminates the need for traditional air-cooling systems that requires water and provides a more flexible and efficient cooling solution.
• Tank/tub immersion cooling: Immersion cooling involves submerging electronic components in a non-conductive liquid that remains liquid at all times. The coolant directly absorbs heat from the components and is then pumped to a heat exchanger. Despite its efficiency, this method has drawbacks such as high costs, careful handling requirements of the hazardous liquid, and accommodation challenges in legacy spaces.
• Direct-to-chip water cooling: This method circulates a liquid directly over the heat-generating components of the server. A cold plate made of a conductive material like copper is attached to the components, allowing the coolant to flow through and pick up heat. The heated coolant is then cooled down again in an external heat exchanger or removed from the system.
• Air-conditioned cooling cabinets: These self-contained units act as micro-data centers, enclosing the equipment and using an air conditioning unit to cool the air within the enclosure. These cabinets help keep equipment at optimal operating temperatures, reducing the risk of overheating and enhancing reliability.

Challenges and considerations

However, edge sustainability strategies can come with several challenges which edge operators should carefully consider:

• High initial costs: Implementing cooling systems can involve significant upfront investment. The costs associated with purchasing and installing specialised cooling equipment can be a barrier for some operators.
• Maintenance requirements: Regular maintenance is essential to ensure the reliability and efficiency of cooling systems. This includes monitoring for leaks, maintaining coolant quality, and servicing components such as pumps and heat exchangers.
• Complexity of retrofitting: Retrofitting existing edge locations with cooling systems can be complex and costly. It may require significant modifications to accommodate new cooling infrastructure, which can be disruptive to ongoing operations.

Tailoring sustainability strategies to different edges

The diverse nature of edge computing environments calls for tailored sustainability strategies. Factors like existing infrastructure, operating environment, and specific needs influence the choice of sustainable solutions. For example, extreme edge solutions, such as those mounted on top of offshore wind turbines will require different cooling solutions than on-premise edge deployments. As such, edge operators need to identify their specific requirements and choose a solution that effectively meets these needs while keeping the total cost of ownership low.

Conclusion

Prioritising sustainability in edge computing is essential for long-term viability and environmental stewardship. By adopting tailored strategies, edge operators can enhance efficiency, reduce costs, and minimise their carbon footprint, contributing to a greener future for cloud and edge computing.