GlobalFoundries says it will spend ~$1B on expanding its largest chip fab, which is in Malta, NY, and create a new fab in Malta via a private-public partnership
Context & Ripple Effects
This July 2021 announcement caps a two-week stretch in which GlobalFoundries laid out its post-dividend-era growth plan: days earlier it had committed over $4B to a $4B+ Singapore plant alongside $1B each for its German and US sites. The Malta expansion — roughly $1B into its largest fab, plus a second Malta fab structured as a private-public partnership — was the first concrete signal that US capacity would anchor that plan.
The partnership framing proved prescient. Three years later the structure became policy reality when the US awarded GlobalFoundries $1.5B in CHIPS Act grants and $1.6B in loans specifically for a new Malta fab and a Fab 8 expansion — and by 2025 the company had scaled its domestic commitment to $13B across New York and Vermont within a $16B program. This 2021 move is where that arc starts.
First-order effects
- GlobalFoundries adds wafer capacity at Fab 8, whose three-month, 700-step production process the Washington Post detailed just weeks earlier, and commits to greenfield construction in Malta rather than only brownfield expansion.
- By structuring the new fab as a private-public partnership, GlobalFoundries shifts part of the capital burden onto government partners — a template it would later monetize directly through CHIPS Act grants and loans.
Second-order effects
- The US commitment gives GlobalFoundries a funded counterweight to its international bets: the $4B Singapore plant opened in 2023 targeting 450K wafers per year, meaning customers now choose between a subsidized US footprint and an Asian one.
- Rival foundry projects face a raised bar — with the joint GlobalFoundries–STMicro effort in France reported stalled while the US-backed Malta path advanced, public co-funding becomes a competitive differentiator among fabs.
Third-order effects
- If the pattern holds, mature-node fab economics restructure around state co-investment: capacity decisions follow subsidy geography first, with GlobalFoundries' later silicon-photonics R&D award extending the model from building fabs to funding the research attached to them.
The trend: Semiconductor capacity is shifting toward publicly co-financed regional clusters, with government partnership moving from exception to default in fab siting decisions.