Published on February 15, 2024, 8:09 pm

“The Growing Challenges Of Power And Storage In The Global Data Center Industry”

The global data center industry is facing significant challenges as power and storage requirements continue to grow exponentially, according to JLL’s data centers global outlook report for 2024. The report highlights the impact of generative AI on data centers, which requires a massive amount of power and exacerbates the scarcity of data center colocation supply caused by regional power limitations.

Generative AI is expected to contribute significantly to the rising global electricity consumption in the coming years. The European Commission estimates that electricity consumption will increase by 60% by 2030. This rise in electricity consumption, driven by high-performance computing (HPC) and AI workloads, puts pressure on data center infrastructure.

Data center operators are now grappling with meeting power requirements ranging from 300 megawatts (MW) to 500MW. This drastic shift in site selection criteria and day-to-day management is challenging for operators. The growing demand for generative AI, especially with the increasing focus on deep learning, has led to a need for more densely clustered and performance-intensive infrastructure compared to traditional data centers.

These specialized infrastructure requirements pose additional challenges in terms of heat management. High computational power generates a considerable amount of heat, requiring advanced cooling systems that are both efficient and cost-effective. The JLL report emphasizes that robust power infrastructure investments and efficient cooling solutions are crucial for preventing hardware damage and maintaining performance.

One of the major concerns highlighted in the report is the strain on global data infrastructure. Ireland’s electricity company EirGrid predicts that electricity demand from data centers could double to 30% of all consumption by 2028, while Denmark expects data center electricity usage to grow from 1% to 15% of total consumption by 2030. Moreover, one-third of Europe’s grid structure is over 40 years old and urgently requires modernization funding.

As total data generation is expected to double in the next five years, reaching 21ZB by 2027, data center capacity needs to reflect this growth. Data center development decisions will be influenced by the management, access, and storage of this enormous volume of data driven by generative AI. Developers should consider investing in sustainable power solutions and alternative cooling methods like liquid immersion to ensure efficient operations.

However, the JLL report discourages the practice of load shedding as a solution to manage power demands. Load shedding can have detrimental effects on data centers that rely on uninterrupted power supply (UPS) for critical equipment operation.

In conclusion, the global outlook for data centers appears optimistic. However, the industry must address challenges posed by increasing power and storage requirements driven by generative AI. The need for robust infrastructure investment, efficient cooling solutions, and support for sustainable power sources cannot be overlooked. As AI-specialized data centers differ from conventional facilities, careful consideration of site selection and infrastructure design is crucial for long-term success in this rapidly evolving industry.

Share.

Comments are closed.