This audio was created using Microsoft Azure Speech Services
This blog post was written in collaboration with guest blogger, Juan Carlos Londoño, who is Presales Director at Ingenium.
Data center demand is higher than ever. Data-intensive technologies such as artificial intelligence (AI), augmented reality, and video streaming are accelerating the need for compute capacity. This creates a challenge because data centers are simultaneously under pressure to reduce their carbon footprint and meet sustainability goals.
In 2024, regulations will take effect in places such as the Eurozone and California, requiring data centers with a load greater than 100Kw to publicly report carbon emissions and climate-related risk management plans. The goal is to reduce carbon emissions in alignment with local regulations such as the U.N.’s goal of achieving Net Zero emissions by 2050.
However, the drive toward sustainability creates friction with the need to support data-intensive technologies and graphics processing units (GPUs), which require a lot of computer power and drive-up demand for electricity. One way to balance these forces is by using more renewable energy sources, but that option isn’t always available to operators.
Lack of Readiness
Data centers need to track power usage and equipment efficiency to fulfill the reporting requirements of carbon-reduction regulations. However, many data center teams need more tools and procedures to track these metrics. Currently, less than one-third of operators are prepared to report data on the average utilization and energy demand of their servers and storage equipment, according to a 2023 survey by the Uptime Institute. The survey found that nearly half of server-based storage systems are misclassified as servers in the operators’ internal databases.
The inability to provide these metrics may tempt some companies to postpone, delay, or cancel their carbon-reduction plans. Some may opt to downgrade their objectives to resolve a potential conflict between fulfilling their sustainability goals or doing the minimum required by the regulatory mandates.
AI Power Demand
The demand for power created by generative AI and GPUs poses a significant challenge by boosting electrical consumption at the rack level. This puts the onus on operators to develop new architectures, design approaches, and more efficient ways to manage cooling, which accounts for 40% of power consumption at data centers. Liquid cooling solutions are likely to become more prevalent.
Increased demand for electricity also requires better visibility across the IT infrastructure to track utilization, power consumption, and overall equipment performance. Operators can achieve this visibility through monitoring solutions that gather and analyze data from data center equipment and the environment around it. Since many operators lack these capabilities, they need to invest in monitoring solutions to better position them to meet sustainability goals.
Demand for Renewables
One obvious solution to resolving the conflict between attaining sustainability goals and supporting power needs is increasing reliance on renewable energy sources. However, the option isn’t always readily available. For example in regions such as Latin America options are limited, even though countries like Costa Rica and Chile are making strides.
Where possible, operators will seek to expand capacity in locations supported by wind, solar, geothermal, and hydroelectric power. Ironically, however, adequate monitoring of utilization and efficiency metrics, which most operators currently lack, could be more significant than using renewables. By tracking metrics, operators can understand utilization and make decisions to turn off resources when they are not needed.
Find the Right Partners
To meet the challenges data centers are facing, operators should work with partners that deliver expert guidance and solutions aimed at meeting sustainability goals. Partners can share best practices on various operational aspects. For instance, when it comes to cooling, it’s possible to run data centers at slightly higher temperatures than most operators realize without risking equipment harm.
The right partners can also steer operators toward equipment designed to be more efficient. These types of investments are crucial to building facilities that can handle higher demand for compute power and electricity while still striving toward sustainability goals. Data centers will not have time to waste as new carbon-reduction regulations go into effect. The time to invest in sustainability and key partnerships is now.
Schneider Electric is a global leader in sustainable data center technology, providing adaptive, resilient, and efficient solutions. We partner with data center operators for complete lifecycle solutions, from design and construction to providing operating and maintenance services. Discover data center solutions from Schneider.
Ingenium is the leading firm in consulting, engineering, construction and operation of data centers, mission-critical infrastructure and facilities in Latin America. We have the knowhow and partner relationships to deliver speed to market for operators to make them more competitive. Learn more about Ingenium.
Add a comment