AI Data Center Boom Needs More Than Chips

The explosive growth of artificial intelligence is not just about cutting-edge semiconductors and advanced software; it's also fueling a massive demand for fundamental building materials. Bloomberg Businessweek's Carol Masser and Tim Stenovec discussed the less-heralded but crucial infrastructure needs of the AI revolution with Katherine Thompson, founding partner and CEO at Thompson Research Group. Thompson, an expert in industrial and construction sectors, shed light on how the construction of AI data centers is spurring growth in traditional industries.

The Unseen Foundation of AI

The AI boom is estimated to be a nearly $1 trillion wave of capital spending, with a significant portion directed towards data center construction. Recent reports indicate that US data center construction spending alone eclipsed $50 billion in April. While semiconductors are a major component, Thompson emphasized that the physical build-out requires a broad range of materials.

"Since the end of World War II, a lot of the focus has been on the consumer leading the economy, but we're having a total flip-flop with this now, and it's re-industrialization," Thompson explained. She noted that the scale of this build-out will eclipse the total spend on the interstate highway system, electrification, and early rail development in the US.

The full discussion can be found on Bloomberg Podcast's YouTube channel.

Beyond Semiconductors: The Essential Materials

Thompson detailed the critical materials needed before any high-tech equipment is even installed in a data center. These include:

• Lime: Essential for stabilizing soil, a crucial step before any foundation can be laid.
• Water Infrastructure: Requiring pre-cast pipes.
• Cement and Rock: For laying foundations and general construction.
• Steel: Forming the structural backbone of the buildings.

"Before you even see any rack, you're dealing with a lot of really basic materials," Thompson stated, highlighting the often-overlooked industrial components driving the AI infrastructure build.

Supply Chain Resilience and Regional Strengths

The question of whether the supply chain can meet this demand is paramount. Thompson pointed out that the availability of these basic materials can vary significantly by region. States with robust production of materials like cement and rock, such as Texas, are better positioned to support the surge in data center construction.

"The areas where it gets it can get tougher um can be more the the rare earth or rare metals that are going to chips. But as far as steel we feel good about um cement could get tight but oddly enough it'll get tight if consumer if if residential returns but right now we seem to be okay," Thompson commented. She cautioned that a recovery in the residential construction market could strain cement supplies.

Investment Opportunities in the AI Build

The conversation then turned to companies poised to benefit from this demand. Thompson identified several key players:

• Commercial Metals Company (NASDAQ:CMC): A steel producer that has transformed by offering construction solutions, including grid products for soil stabilization and rebar for concrete foundations. The company recently outlined solutions for a $17 billion semiconductor plant buildout.
• Worthington Enterprises: Provides large metal tanks for gases and essential construction materials like steel studs and ceiling tiles, as well as grid products.
• Martin Marietta Materials: A major producer of crushed rock and lime. The company is in the process of acquiring the largest lime producer in the US, positioning it to capitalize on the demand for soil stabilization materials.

Thompson noted that while semiconductors are the headline-grabbing aspect of AI, the underlying industrial and construction sectors are experiencing significant growth. She observed that the steel index, for instance, is up nearly 45% year-to-date, alongside gains in heavy electronics and construction and engineering sub-indices.

"This is a market where we see rotation where it's not all about the sexy names," she concluded, suggesting that investors look beyond the obvious tech giants to find value in the essential industries supporting the AI revolution.