Shares of Intel surged 7.7% to ARS 34,780 on September 18 as the market digested reports that SK Hynix, the world's dominant supplier of high-bandwidth memory used in AI chips, is exploring a deal to manufacture memory at Intel's massive Ohio campus. The talks remain exploratory, with no decisions made , yet the stock's move reflects investor hunger for any evidence that Intel's foundry ambitions are attracting real tenants.
• A $100 Billion Campus Finally Has a Potential Customer. Intel's Ohio site spans roughly 1,000 acres — large enough for eight semiconductor plants — with a full buildout estimated at $100 billion.
The two planned fabs are now not expected to finish until 2030 and 2031 , years behind the original 2025 target. Leasing idle capacity to SK Hynix would turn a cash drain into rental income and help justify the billions already poured into concrete and steel.
• Intel's Foundry Still Barely Has Outside Revenue. Intel Foundry booked $5.8 billion of revenue in Q2 2026, but only $293 million came from external customers.
For full-year 2025, the foundry posted a $10.3 billion operating loss. Landing SK Hynix — even as a tenant, not a traditional foundry client — would be Intel's most significant external validation yet. But lease revenue is not the same as high-margin chip manufacturing revenue, and the financial terms remain undefined.
• SK Hynix Has Leverage — and Competing Options. SK Hynix broke ground on a $4 billion-plus advanced packaging plant in Indiana in late August, with mass production targeted for late 2029.
The company commands roughly 58% of worldwide high-bandwidth memory revenue. It can afford to negotiate hard; Intel, burning cash on a foundry that primarily serves itself, arguably cannot.
• The Stock Has Run Far Ahead of Foundry Fundamentals. Intel shares have surged more than five-fold in the past year, from roughly $19 to around $120.
The risk is that Intel is spending over $20 billion this year building capacity whose external buyer has not yet signed. A preliminary conversation — no matter how strategically logical — does not guarantee a binding contract, equipment orders, or revenue. No fabrication equipment has been ordered, a step that itself carries roughly a two-year lead time.
The bottom line: the SK Hynix talks validate Intel's foundry-landlord strategy in principle, but the gap between exploratory dialogue and signed revenue remains enormous — and the stock price already assumes the best outcome.