The bottleneck on AI data centers isn't semiconductors. It's a machine with a three-year lead time
In June I wrote a piece called "Investment outlook for the second half of 2026." Middle East risk versus the AI trade. Three scenarios. Seven things to watch. It got zero likes. I understand why now. There wasn't a single line in it that only I could have written. The oil price outlook, the Fed's stance, the P/E on the Nikkei — I had carefully rearranged that day's news. The numbers were borrowed…
Three years separates the construction of AI data centers from the delivery of the machines that power them. Power plants don't arrive immediately when needed; they require lead times of three to five years for the equipment that generates, transforms, and delivers electricity. LNG supply contracts extend even further, typically spanning ten to twenty years.
This mismatch in timeframes creates a bottleneck for the growth of AI infrastructure. Companies are forced to secure equipment and negotiate contracts well in advance, often before the project's commercial operation date is even set. The result is a shortage of delivery dates for power equipment, not a shortage of electricity itself.
The implication is that AI growth is constrained not by the availability of power, but by the time it takes to procure and deploy the necessary machines.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
