There is a growing focus on datacentre energy consumption and operational efficiency. Legislators and regulators are taking steps to require operators to report facility-level information and key performance indicators to control datacentre growth. The aim is to establish minimum facility performance thresholds and require operators to report indicative performance and efficiency metrics such as power usage effectiveness (PUE) and work delivered per unit of energy consumed.
The European Union has taken the lead in this effort with the Energy Efficiency Directive (EED) and a finalized but unpublished Delegated regulation. These regulations mandate that datacentres with over 500 kilowatts of installed IT equipment power demand report more than 30 location and operating parameters to member states and the European database on datacentres. The final delegated regulation expands reporting requirements to include installed server work capacity and installed storage capacity.
Even without regulatory mandates, utilizing a work per energy metric is considered a best practice in the industry. Each new generation of IT equipment offers significantly higher work capacities and work per watt compared to previous generations. Advancements in workload management software also enable operators to maximize the utilization of their IT equipment. A work per energy metric effectively demonstrates the benefits of refreshing and consolidating IT equipment.
However, datacentre operators face challenges in reporting server and storage capacity indicators. To comply with regulatory requirements or calculate work per energy metrics, operators need to maintain an equipment inventory with critical component data and location information. They must also establish processes to capture and calculate server work capacity. The industry requires a standardized method for reporting server work capacity.
According to a survey, only one-third of IT operators maintain a detailed equipment inventory capable of calculating equipment capacities in a data centre. Additionally, only 30% of operators can match inventory data with the quantity and type of equipment in each facility. Furthermore, only 27% of operators collect CPU part numbers and core counts for calculating server work capacity, while 53% collect data on storage capacity.
To improve their equipment inventory and management processes, datacentre operators need to update their inventory systems to include all the necessary component information for calculating work capacity. They should also update equipment purchase specifications to require the reporting and collection of component data. Conducting a survey of the installed IT equipment can provide a complete inventory, either through asset discovery software or a manual survey.
Building an effective inventory system will require collaboration among different departments and will take time. However, 68% of operators believe they can implement a system within one year.
The delegated regulation defines server work capacity as the active state performance score designated by the Server Efficiency Rating Tool (SERT®). The majority of the datacentre industry supports using the SERT score as a representative server work capacity value. The Green Grid (TGG) has collected a dataset of SERT measurements for over 600 server configurations, which will be used to establish databases containing active state performance values for CPU part numbers.
For storage equipment capacity, the work capacity is simply the raw storage capacity measured in terabytes, readily available from the product manufacturer. The value of a datacentre’s storage capacity is the sum of the raw storage capacities of all installed storage products.
In conclusion, datacentre operators and the industry as a whole need to work towards building the necessary datasets and databases to calculate work capacity for regulatory reporting. This includes improving equipment inventory systems, collecting component data, and collaborating with manufacturers and industry organizations.
