Consultants Forum Advertise
Subscribe
Close

  • Free for qualified executives and consultants to industry

  • Receive quarterly issues of Area Development Magazine and special market report and directory issues

Subscribe Now
Renew

Area Development - Site Selection, Facility Planning and Workforce Development

In Focus: AI’s Influence on Data Center Growth

The use of AI is driving a surge in demand for computing power and the development of more data centers, which, in turn, have an increasing need for reliable power and water.

Q4 2023
Artificial intelligence (AI) is one of the most talked about subjects of the year, anticipated to touch almost all facets of modern life. Hotly debated from dinner tables to Hollywood to the halls of Congress, synthetic data will completely overtake real data in AI models by 2030, according to market research firm Gartner.

The increasing adoption and emphasis on AI as a core component of corporate objectives is driving a substantial surge in demand for computing power and resources required to implement AI models. This new demand requires the development of more — and better — data centers that can support the unprecedented computing power required to operate AI.

The AI boom is coming at a time when the data center market is already experiencing increased demand. According to the 2023 JLL Global Data Center Outlook, the global colocation data center market is forecast to grow at 11.3 percent per annum (CAGR) from 2021 to 2026. We expect the hyperscale data center market to grow even faster at an approximately 20 percent CAGR. The rise of AI will both bolster the already high demand — with increased data usage — and improve computing efficiency.

The AI boom is coming at a time when the data center market is already experiencing increased demand. Location Factors
From power to air cooling requirements, legacy data centers are generally unequipped for AI. Faced with either upgrading existing facilities or developing new centers, operators and occupiers must make hard choices. In preparation for an AI-driven future, it is crucial to approach site selection with a fresh perspective. Exploring new submarkets may reveal the ideal solution, while considering proximity needs becomes a significant decision-making factor.

When it comes to universally applicable AI tools like ChatGPT, the location of a data center can be more adaptable; however, for financial services firms, retailers, and other organizations using proprietary, industry-specific AI models, being near existing energy grids, connectivity infrastructure, and network hubs is essential.

Ensuring a reliable power supply is paramount, and innovative solutions can be uncovered in nontraditional settings. Regardless of the chosen location, ensuring a reliable power supply is paramount, and innovative solutions can be uncovered in nontraditional settings. For new centers in these emerging markets, data center operators need to install infrastructure to accommodate high power density server clusters to accommodate AI requirements. Keep in mind that AI needs as much as five times the fiber, so this may lead to an infrastructure boom to AI sites.

AI also needs more power and ability to keep the servers cool. Unfortunately, AI’s hunger for power and water are at odds with current ESG goals and decarbonization efforts. Innovation is crucial to improve cooling and energy efficiency for AI uses given sustainability goals of hyperscalers and colocation providers.

With the continuous reduction in transistor sizes to nanometer scale and beyond, the adoption of liquid cooling solutions emerges as a popular solution, as it allows data centers to meet sustainability goals and reduce carbon footprints. Other options include using solar panels and other renewable energy sources. While certain questions remain unanswered, we cannot overstate the importance of addressing both power demands and thermal requirements for chips when it comes to the design of future data centers.

Exclusive Research