TL;DR — Intel’s second-quarter revenue rose 25% to $16.1 billion, with Data Center and AI revenue up 59% to $6.3 billion, its strongest growth in over 15 years. It still posted an $11.0 billion GAAP net loss, and $13.6 billion of that came from a non-cash charge: Intel’s stock rose so much that the value of shares it owes the U.S. government under its CHIPS Act deal increased, and Intel had to book that increase as a loss.
Intel reported second-quarter revenue of $16.1 billion, up 25% year over year — its strongest growth rate in more than fifteen years. Data Center and AI revenue jumped 59% to $6.3 billion. By every operating measure, this was Intel’s best quarter in years. The headline number was still a net loss of $11.0 billion, wider than the $2.9 billion loss a year earlier. The loss didn’t come from the business getting worse. It came from Intel’s stock getting better.
The Mechanism Behind an $11 Billion Loss
As part of its CHIPS Act “Secure Enclave” agreement with the U.S. Department of Commerce, Intel placed shares in escrow that convert to government ownership as funding is disbursed. As of the end of the quarter, Intel had released just 13 million of 143 million unreleased escrowed shares. Intel carries the unreleased balance as a derivative liability on its balance sheet, revalued every quarter to the current stock price. When the stock rises, the liability Intel owes gets more valuable — and Intel has to record that increase as a loss, even though no cash changes hands and no shares have actually been transferred.
That mark-to-market charge came to $13.6 billion for the first half of 2026. Back it out, and Intel’s underlying quarter was solidly profitable. The company’s own stock rally — driven by the CHIPS Act deal itself, plus separate investments from Nvidia and SoftBank earlier this year — is the direct cause of the accounting loss investors saw on the headline.
Intel’s stock did well enough this quarter to make its earnings look terrible. That’s not a paradox — it’s just how a $13.6 billion derivative liability works.
The Loss That Wasn’t an Accounting Trick
Not every part of Intel’s loss column is a non-cash artifact. Intel Foundry, the division building out contract chip manufacturing to compete with TSMC’s advanced packaging capacity, posted a real operating loss of $2.089 billion on $5.8 billion of revenue, up 31% year over year. That’s a genuine cash-burning loss, not a mark-to-market entry. The better news: it’s narrower than the $3.2 billion Foundry lost in the same quarter last year, an improvement of roughly 36%, and operating margin moved from -71.7% to -36.2%.
- Total company: $16.1B revenue (+25%), $11.0B GAAP net loss, driven mostly by a non-cash escrow charge.
- Data Center and AI: $6.3B revenue, up 59% — Intel’s fastest-growing segment.
- Intel Foundry: $5.8B revenue (+31%), -$2.1B operating loss, narrowing from -$3.2B a year ago.
Foundry still isn’t profitable, and won’t be for a while — Intel is targeting high-volume manufacturing on its 14A node by 2028. But a loss that’s shrinking because of tightening costs and improving yields is a fundamentally different story than a loss caused by a government equity stake appreciating in value. Investors and headline writers alike tend to collapse both into the same $11 billion number.
What the Market Actually Did
Intel shares initially surged on the earnings report before retreating to around $100 in the days after, as investors worked through which parts of the loss were real. For the third quarter, Intel guided to revenue of $15.8 billion to $16.8 billion, with GAAP EPS of $0.31 and non-GAAP EPS of $0.38 — both positive, and both suggesting management doesn’t expect another escrow-driven swing to erase the quarter’s results the way this one did.
The number to watch next quarter isn’t the GAAP net income line. It’s whether Data Center and AI revenue keeps growing at anywhere near 59%, and whether Foundry’s loss keeps narrowing at the same pace it did this quarter.
Figures are from Intel’s second-quarter 2026 earnings release and its Form 10-Q filed with SEC EDGAR, for the quarter ended June 27, 2026, with additional detail on the Foundry margin trend from FourWeekMBA’s analysis of the results. Third Pole Markets holds no position in Intel as of publication — see our About page for the full disclosure policy.
