In a strategy shift, GlobalFoundries announces it is ceasing all work on its 7nm fabrication processes
Context & Ripple Effects
GlobalFoundries' 7nm exit is the pivot point that defines everything the company has done since. Spun out of AMD in 2009, it spent years chasing the leading edge before concluding the capital cost was untenable — and the later coverage confirms the bet: by 2023 it was positioning itself as a maker of 'essential' 12nm-and-above chips for image sensors and power management, per a company profile, and in 2021 it committed roughly $1B to expanding its Malta, NY fab rather than building toward smaller nodes (Malta expansion).
The move also foreshadowed an industry-wide reckoning with leading-edge economics. In 2025, Reuters reported Intel considering writing off its 18A processes to concentrate on 14A against TSMC — evidence that the cost wall GlobalFoundries hit in 2018 still stands, now with fewer players left on the far side of it.
First-order effects
- GlobalFoundries halts its most advanced roadmap, redirecting engineering and capital toward mature nodes where its Malta fab anchors production — and any customer needing 7nm or below must source elsewhere, effectively feeding volume to TSMC.
Second-order effects
- With GlobalFoundries out, the leading-edge foundry field thins, strengthening TSMC's pricing and allocation leverage — leverage it later exercised directly when it told Chinese customers it would stop manufacturing AI chips at 7nm and below (TSMC's 7nm cutoff).
Third-order effects
- If the pattern holds, the industry settles into a two-tier structure: one or two firms able to fund successive leading-edge nodes — a bar Intel's own 18A struggles illustrate — and a cohort of specialists like GlobalFoundries competing on mature-node capacity, cost, and supply security.
The trend: Leading-edge chipmaking is consolidating into an ever-smaller set of foundries while the rest of the industry repositions around mature, 'essential' process nodes.