The race for AI computing power just hit a new milestone. A major semiconductor manufacturer is ramping up capacity by constructing a new fabrication facility in New York, responding to skyrocketing demand for high-bandwidth memory chips powering AI systems.

The numbers are staggering: $100 billion investment to build out this operation. But here's the catch—the facility won't reach full production capacity until 2041. That's a 16-year runway, underlining both the massive scale of infrastructure required and the long-term confidence betting on sustained AI chip demand.

This expansion reflects a broader shift in how computing infrastructure is scaling. Whether it's AI workloads, blockchain validation networks, or data centers, chipmaker capacity is becoming a critical bottleneck. The extended timeline also signals that even with aggressive investment, supply constraints will likely persist across the industry for years to come.