Copper demand from data centers through 2030 is on a trajectory that the global mining industry cannot physically satisfy — and the arithmetic is straightforward enough that any investor willing to do the math should be structurally positioned in copper right now.
A single hyperscale data center campus — the kind being planned by Microsoft, Google, Amazon, and Meta across the United States — requires approximately 50,000 tonnes of copper just to build. Wiring, transformers, busbars, cooling systems, power distribution — copper is the circulatory system of every data center on earth. The United States is planning 13 to 14 campus-scale facilities. That is 650,000 to 700,000 tonnes of copper demand from data centers alone, before a single EV is manufactured or a single grid upgrade is completed.
Total global copper mine production runs at approximately 22 million tonnes per year. The data center buildout alone represents more than 3% of annual global supply concentrated into a multi-year construction window, competing with electrification, defense manufacturing, and consumer electronics for the same constrained supply.
Craig Tindale’s point in his Financial Sense interview bears repeating: a copper mine takes 19 years from discovery to full production. Robert Friedland just brought one of the world’s largest new copper mines online in the DRC, and Tindale’s analysis suggests we would need five or six mines of equivalent scale opening every year just to keep pace with demand growth through 2030. We are not opening five or six. We are opening one.
The copper demand data centers 2030 story is not a commodity cycle. It is a structural supply deficit driven by the physical requirements of the infrastructure the technology industry has already committed to building. That deficit will be priced — the question is whether you’re in front of it or behind it.
The Hedge 