A pledge that keeps growing every time Micron speaks

Micron Technology announced Thursday, July 9, that it now plans to invest more than $250 billion in U.S. semiconductor manufacturing through 2035 — a $50 billion jump from the $200 billion commitment the company made just last June, which itself represented a $30 billion increase from Micron's original spending plans. That's three upward revisions to the same long-term investment figure inside roughly a year, each one arriving alongside continued strength in AI-driven demand for the memory chips Micron manufactures.

Investors responded immediately. Micron shares rose about 8% in early trading following the announcement, according to Reuters, adding to a stock that has already surged more than 200% so far in 2026. That single-day pop reflects less a surprise reaction to the headline number and more a market that's grown accustomed to reading upward Micron capex revisions as a leading indicator of just how tight AI memory chip demand actually is.

What's driving the number higher each time

Sanjay Mehrotra, Micron's chairman, president, and CEO, tied the timing explicitly to a broader national milestone in his statement announcing the increase: "As America celebrates its 250th anniversary, data and memory are foundational to the modern economy — and Micron is increasing our US investments to more than $250 billion through 2035 to meet that moment." He specifically thanked President Trump, Commerce Secretary Howard Lutnick, New York Governor Kathy Hochul, Senator Chuck Schumer, and local Central New York officials for what he described as their leadership in supporting the buildout — a notably bipartisan list of political credit for a company navigating both federal semiconductor policy priorities and state-level economic development incentives simultaneously.

The underlying demand driver isn't abstract. Micron disclosed last month that customers across the data center, consumer electronics, and automotive sectors had already locked in roughly $22 billion worth of memory chip orders through the company's Strategic Customer Agreements — long-term, binding contracts that give Micron unusually strong visibility into future revenue relative to typical semiconductor demand cycles, which are notoriously difficult to forecast years in advance.

The GlobalWafers deal: securing the raw material, not just the factory

Buried inside the headline $250 billion figure is a smaller but strategically important piece: Micron will direct up to $3 billion toward strengthening the broader U.S. semiconductor supply chain, with $500 million of that specifically earmarked for GlobalWafers, a Taiwan-based silicon wafer manufacturer, to expand its 300-millimeter raw silicon wafer facility in Sherman, Texas. The two companies will simultaneously enter a 10-year supply agreement giving Micron guaranteed access to what the companies describe as "significant" raw silicon wafer capacity, according to Reuters and Data Center Dynamics' reporting on the deal.

This detail matters because it addresses a different kind of bottleneck than fab construction alone. Building advanced memory chip factories solves for manufacturing capacity, but every one of those factories still needs a steady, secured supply of raw silicon wafers as the base material chips are actually built on. Locking in a decade-long wafer supply agreement, with direct financial backing to help expand that supplier's own manufacturing capacity, is Micron effectively vertically securing its upstream materials pipeline at the same time it's building out downstream fab capacity — a hedge against the kind of supply chain disruption that rattled the broader chip industry during the pandemic era.

New York's fab timeline, from groundbreaking to first output

The centerpiece of Micron's expanded U.S. footprint remains its semiconductor campus in Central New York, where the company has now begun concrete pouring for the first fabrication plant, according to TrendForce's reporting. Per Micron's own published schedule, that first fab is expected to be equipment-ready by the second quarter of 2028, with operations beginning in early 2029. Three additional fabs at the same campus are slated to come online progressively — the second by late 2030, the third by 2035, and a fourth stretching out to 2041, giving the New York site a genuinely multi-decade buildout horizon rather than a single, near-term construction project.

That timeline isn't limited to New York. Micron confirmed construction is also underway on two new fabs in Boise, Idaho, with first wafer output expected in mid-2027 for the initial facility and late 2028 for the second, while a Virginia facility already began initial production earlier this year. Spreading fab construction across three separate states, on staggered but overlapping timelines, gives Micron a production ramp that isn't dependent on any single site's construction schedule running smoothly — a meaningful risk mitigation given how often large-scale fab construction projects, across the industry generally, run behind their original schedules.

A broader industry pattern, not an isolated bet

Micron's expanded commitment lands within a considerably larger wave of announced domestic chip investment. Competitors Samsung Electronics and SK Hynix have separately announced plans to invest up to a combined $880 billion in new chip production facilities, according to Times Now World's coverage — figures that, taken together with Micron's own $250 billion pledge, suggest the entire memory chip industry is making a coordinated, multi-decade bet that AI-driven demand for high-bandwidth memory and related products isn't a temporary spike, but a durable, structural shift in how much memory the global computing infrastructure actually needs.

That's the framing worth holding onto here. A single company raising its own investment target by $50 billion is notable on its own. Three of the industry's largest players collectively committing more than a trillion dollars to expanded chip manufacturing capacity, largely concentrated in the U.S. and largely justified by AI infrastructure demand, is a considerably bigger signal about where these companies believe the next decade of computing demand is actually headed — and how much physical manufacturing capacity they believe will be required to meet it.

*This article was researched using publicly available reporting from Reuters, Data Center Dynamics, TrendForce, TelecomLead, Times Now World, and Yahoo Finance coverage of Micron Technology's expanded U.S. investment announcement. It is intended for informational purposes and does not constitute financial advice.*