What was changed
GlobalFoundries was created from AMD's manufacturing assets and backed by Abu Dhabi's Mubadala, but competing with TSMC at the leading edge required vast capital. In 2018 the company put 7nm development on hold indefinitely and chose not to chase the smallest nodes.
The decision looked like a retreat because AMD moved 7nm production to TSMC and the foundry race narrowed. The new strategy was to invest in differentiated offerings on mature and specialty processes, including RF, embedded memory, low power, automotive, IoT, and 5G-related chips.
The public-market result came in 2021. Reuters reported GlobalFoundries was aiming for about a $25bn IPO valuation, and the company later reported FY2021 revenue of $6.6bn, up 36%, with 30 long-term customer agreements committing more than $3.2bn.
What it achieved
By 2021, revenue rose 36% to $6.6bn and the company began Nasdaq trading after seeking roughly a $25bn IPO valuation.
Why it worked
The 7nm race favored a few scale leaders; GlobalFoundries had only about 10% foundry share in 2017.
Specialty processes matched demand from automotive, RF, 5G, and industrial customers that did not always need the newest node.
Mubadala's ownership allowed a painful portfolio reset before public investors judged the business.
What can be applied
A technology retreat can be a strategic advance if it exits a losing capital race and concentrates investment where customers still pay.
Aftermath
GlobalFoundries began trading on Nasdaq under GFS in October 2021. It remained behind TSMC in leading-edge logic, but its repositioning made it a strategically important supplier in a capacity-constrained chip market rather than a subscale pursuer of every new node.
FOLLOW THE EVIDENCE