TSMC is accelerating its Arizona factory buildout. CFO Wendell Huang told CNBC this comes from robust customer demand. Not weak demand. Not tepid demand. Robust.
The company wants to capitalize on the AI megatrend. That's what Huang called it. A megatrend. Not a trend. Not a fad. A megatrend. Like how Pet Rocks were a megatrend in 1975, except this time it's semiconductors and everyone's pretending the capex cycle won't eventually murder them.
Arizona gets the factory because TSMC looked at a map and said, "You know what this desert needs? Clean rooms." The state that brings you 115-degree summers and scorpions in your mailbox now gets to host the most technically demanding manufacturing process on Earth. Someone in Phoenix is about to learn what humidity control costs when ambient temperature wants you dead.
Retail traders saw this headline and immediately checked if TSMC stock moved. It didn't. They checked if Nvidia moved. It did, but for unrelated reasons. They checked if their portfolio moved. It moved down, because that's what it does.
Huang cited customer demand as the driver. Customers demanded chips. TSMC responded by building a factory that won't produce meaningful volume until 2025 at the earliest. This is how you capitalize on a megatrend: slowly, expensively, and in a place where water is a theoretical concept.
The AI boom needs chips. TSMC makes chips. Arizona needs jobs. TSMC needs subsidies. Everyone gets what they want except the guy who bought TSMC calls last Tuesday because he read "megatrend" and thought it meant something.
The buildout accelerates. The timeline compresses. The capex expands. And somewhere in Scottsdale, a day trader just learned that factories take years to build and his options expire Friday.
Photo by on Unsplash
Leave a Comment