Micon’s $100 billion bet that won’t pay off until 2041

Micron Technology (MU) has announced plans to build a massive semiconductor manufacturing campus in New York, aiming to secure a larger share of the memory chips powering the artificial intelligence boom.

The caveat is that the project’s scale comes with a long runway: the full buildout is expected to take decades, underscoring that Micron is positioning for long-term AI demand rather than offering a quick fix to today’s chip shortages.

On Jan. 7, Micron unveiled plans for a roughly $100 billion investment in central New York, a multistage project that could stretch into the early 2040s, with full completion targeted around 2041.

The company said the facility aligns with the Trump administration’s push to reindustrialize the U.S. economy, describing it as a future home to “the most advanced memory manufacturing in the world.”

AI has triggered explosive demand for high-bandwidth memory, a critical component for training and running large-scale models.

However, the infrastructure required to produce advanced memory chips operates on decades-long timelines. Even with enormous capital commitments, semiconductor fabs take years to permit, build, equip, and ramp to full production.

As a result, AI hardware bottlenecks, particularly in memory, are likely to persist through much of the decade.

That dynamic ultimately favors Micron. As one of the few U.S.-based firms capable of producing cutting-edge memory at scale, the company stands to benefit from sustained pricing power, long-term supply agreements, and government-backed incentives.

Micron’s DRAM edge is fueling a market-beating rally

While companies like Nvidia have captured the spotlight during the AI boom, Micron has quietly underpinned the entire ecosystem through its dominance in DRAM, a high-speed memory essential for training and running large AI models.

Unlike processors, which handle computation, DRAM feeds data to AI accelerators in real time, making memory capacity and bandwidth a critical bottleneck as models scale.

That dynamic helps explain why Micron has dramatically outperformed the broader market, despite trading at far lower valuation multiples than Nvidia or Advanced Micro Devices. As demand for high-bandwidth memory has surged, Micron has benefited from tightening supply and rising prices across the memory market.

For those reasons, analysts at Goldman Sachs expect Micron to sustain elevated margins, especially as near-term competition in advanced memory production remains limited.

The stock is up nearly 250% over the past 12 months, reflecting investor confidence that memory has become one of the most valuable chokepoints in the AI supply chain.