Micron plans $24 billion Singapore chip investment as AI demand tightens memory supply

Micron Technology is ramping up its long-term manufacturing footprint in Asia with a major new investment in Singapore. The U.S. memory-chip maker said on Tuesday, January 27, 2026, that it plans to build a roughly $24 billion advanced wafer fabrication facility over the next decade, a move aimed at meeting surging demand for memory used in AI infrastructure, data centers, and data-heavy consumer devices.

A decade-long build focused on advanced NAND

According to the company, the Singapore project will expand its existing manufacturing complex and support production of NAND flash memory—an essential component for storage across phones, PCs, servers, and cloud systems. Micron said wafer output from the new facility is expected to begin in the second half of 2028, indicating the project is designed for the next cycle of demand rather than a quick capacity release.

The scale underscores how sharply the memory market has changed. AI-related spending has boosted demand not only for GPUs and specialized processors, but also for the memory and storage systems that keep those chips fed with data. That pressure has tightened supply and helped memory pricing and margins rebound, encouraging producers to plan new capacity—while still trying to avoid the oversupply mistakes that have historically punished the sector.

High-bandwidth memory ties and Singapore’s role

Micron’s announcement also pointed to related activity in Singapore tied to high-bandwidth memory (HBM), a category of memory increasingly important for AI accelerators. While the new fab centers on NAND, the broader site’s capabilities and buildout signal Micron’s intent to position Singapore as a key hub in its global supply chain for advanced memory products.

Singapore has already been central to Micron’s flash-memory operations, and the company has highlighted the country’s manufacturing ecosystem, workforce, and logistics advantages. For the broader industry, the investment is another indicator that the AI boom is driving capacity decisions that will shape chip supply well into the late 2020s.

Strategic implications for the memory cycle

The key question for investors is whether the industry can expand supply without recreating the boom-bust dynamic that has historically defined memory. Micron’s timeline—production starting in 2028—suggests discipline, but it also raises a competitive issue: rivals will also invest, and technology transitions can shift market share if execution slips.

For customers building AI data centers, additional committed capacity offers reassurance that memory constraints may ease over time. For Micron, the project is a long bet that AI-driven data growth is structural, not cyclical—and that the company can capture that demand with the right mix of manufacturing scale and advanced process technology.