The AI economy extends far beyond Silicon Valley and the companies making headlines. While OpenAI, Google, Microsoft, and NVIDIA dominate conversations about artificial intelligence, a single chatbot query actually triggers a sprawling industrial supply chain that includes mining operations, electrical equipment manufacturers, construction firms, and power infrastructure companies. According to McKinsey's analysis, approximately $5.2 trillion will be invested in AI infrastructure by 2030, and the majority of this capital will not go to neural network developers at all.

Where Is the $5.2 Trillion in AI Infrastructure Investment Actually Going?

McKinsey breaks down the projected $5.2 trillion in AI infrastructure spending into three major categories, revealing a distribution that surprises most observers:

• Computing Equipment: About $3.1 trillion, or 60%, will fund processors, memory, servers, and other computer hardware that powers AI systems.
• Power and Infrastructure: Approximately $1.3 trillion, or 25%, will go toward power generation, electrical grids, transformers, generators, cooling systems, and telecommunications infrastructure.
• Construction and Real Estate: Roughly $800 billion, or 15%, will cover land acquisition, construction materials, and the physical building and outfitting of data centers themselves.

These figures represent only capital expenditures and do not account for ongoing costs like electricity, water, maintenance, repairs, staffing, and equipment upgrades. This distinction matters because it shows the AI boom is reshaping traditional industries in ways most investors have overlooked.

Why Are Copper Miners and Raw Material Suppliers Becoming Unexpected Beneficiaries?

One of the most striking findings is that copper, a metal known to humanity for thousands of years, has become essential to AI infrastructure. Copper is needed in cables, electric motors, transformers, server power systems, generators, cooling systems, and the high-voltage lines that deliver electricity to data centers. S&P Global explicitly identifies AI as a major new source of demand for copper, forecasting that data centers could increase their share of U.S. electricity consumption from the current 5% to 14% by 2030.

This creates a dual demand surge. First, data centers themselves require massive amounts of copper for internal wiring and power distribution. Second, the surrounding energy infrastructure needed to supply these facilities requires additional copper for transformers, substations, and transmission lines. As a result, mining giants, cable manufacturers, metallurgical companies, and electrical raw material suppliers are becoming indirect but significant beneficiaries of the AI boom. This represents a fundamental shift in technology business logic: scaling modern AI models eventually requires extracting more metal from the ground.

How Is the Semiconductor Manufacturing Supply Chain Capturing AI Investment?

While GPUs and AI accelerators receive the most attention, the companies manufacturing the equipment that produces these chips are experiencing explosive growth. Taiwan's TSMC, one of the world's leading chip manufacturers, saw its high-performance computing segment account for 66% of revenue in the second quarter of 2026, with that segment growing by an additional 20% in the same quarter. The company attributes this strong demand primarily to AI and high-performance computing applications.

However, TSMC itself depends on suppliers further up the chain. The Dutch company ASML manufactures the lithography systems that semiconductor manufacturers use to create cutting-edge chips. In the second quarter of 2026, ASML reported 9.3 billion euros in revenue and 2.9 billion euros in net income, with the company forecasting 43 to 45 billion euros in sales for the full year 2026. ASML directly attributes this surge to large-scale AI investments forcing chip manufacturers to accelerate factory expansions.

"ASML directly attributes the rise in demand to large-scale AI investments, which are forcing logic and memory chip manufacturers to accelerate their factory expansions," noted McKinsey analysts reviewing the semiconductor supply chain.

McKinsey Analysis, Source 1

This creates a revenue pyramid: NVIDIA sells the processor, TSMC earns money from manufacturing it, and ASML profits from the equipment without which TSMC cannot create cutting-edge processes. Simultaneously, manufacturers of chemicals, silicon wafers, gases, memory components, packaging materials, and hundreds of component suppliers all capture revenue from this single AI accelerator.

What Role Do Electrical Equipment and Cooling System Manufacturers Play?

Modern AI data centers consume hundreds of megawatts of electricity, requiring specialized infrastructure that extends far beyond the building itself. Before power reaches the servers, it must flow through substations, transformers, distribution systems, uninterruptible power supplies (UPS) units, and other equipment. Demand for this infrastructure is growing faster than manufacturers can increase production.

McKinsey reports that lead times for certain types of equipment in North America have stretched to approximately 80 weeks for medium-voltage switchgear and 50 weeks for transformers. One of the most telling examples of this new boom is Eaton, a U.S.-Irish company that manufactures electrical equipment for data centers, industry, and the energy sector. In the second quarter of 2026, Eaton's sales reached a record $8.5 billion, up 21% year-over-year.

How Should Investors and Industry Observers Think About AI Infrastructure Economics?

The AI boom increasingly resembles one of the largest infrastructure projects of the 21st century rather than simply a technological revolution. The scale of investment required means that traditional industries, from mining and metallurgy to electrical equipment manufacturing and construction, are capturing substantial portions of AI-related capital spending. Understanding this broader economic picture is essential for investors seeking exposure to AI growth beyond the obvious semiconductor and software companies.

The concentration of spending across multiple industries also suggests that AI infrastructure development will face real-world constraints. Supply chain bottlenecks in transformers, copper availability, and construction capacity could become limiting factors in how quickly data centers can be deployed. Companies and investors focused solely on chip performance may miss the infrastructure constraints that ultimately determine how fast the AI economy can actually scale.