Thursday was the kind of day that makes headlines for the wrong reasons. The Magnificent Seven lost $767 billion in combined market value. Tesla fell 14.3%. Alphabet fell 6.9%. Brent crude crossed $100 a barrel. The 10-year yield hit 4.71%, its highest since January 2025. And buried under all of it, Intel reported its strongest revenue growth in more than 15 years, beat earnings estimates by 91%, and saw its stock jump 10% in afterhours trading. In a session where everything that was supposed to be good turned red, the company that was supposed to be a turnaround story quietly delivered one.
Intel's Quarter: What the Numbers Actually Say
Revenue came in at $16.1 billion, up 25% year over year, beating the $14.45 billion consensus by 12%. Non-GAAP EPS of $0.42 demolished the $0.22 estimate by 91%, up 520% year over year. CEO Lip-Bu Tan called it the company's strongest revenue growth in more than 15 years, and the segment breakdown shows why.
Data Center and AI revenue grew 59% to $6.3 billion. Intel Foundry grew 31% to $5.8 billion. Client Computing and Physical AI grew 13% to $8.9 billion. Intel 18A-P, the advanced manufacturing node that the market had largely given up on, has entered risk production with yields improving meaningfully. The CFO announced plans to raise 2026 capex from $18 billion to $20 billion, signed long-term customer agreements covering CPUs and XPUs with volume and price commitments, and guided Q3 revenue to $15.8 to $16.8 billion, above consensus.
On an operational basis, Intel is a different company from the one that was losing market share, missing process nodes, and reporting GAAP losses that made investors question whether the turnaround was real. Thursday's results make the case that it is real.
The $11 Billion Paper Loss That Confused Everyone
The GAAP result looked alarming: a net loss of $11 billion on a quarter where the business beat every single metric. The explanation is one of the more unusual accounting mechanics in the market right now.
Under Intel's CHIPS Act agreement with the US Department of Commerce, Intel holds shares in escrow for the government. Accounting rules classify those escrowed shares as a derivative liability. When Intel's own stock price increases, the marked-to-market value of that liability rises, and Intel must book the increase as a non-cash loss. Q2's $12.5 billion charge came entirely from this mechanism. Strip it out and Intel earned $2.2 billion in operating income.
In other words, Intel lost $11 billion on paper largely because its own stock became more valuable. The CHIPS Act, which was designed to strengthen US semiconductor manufacturing, created an accounting structure where Intel is penalized on paper for its own share price appreciation. Once this is understood, the GAAP loss is not a concern. It is a reminder to read through GAAP to operating reality for any company participating in CHIPS Act subsidy structures.
Why Intel's AI and Foundry Growth Matters
The 59% growth in Data Center and AI is not a standalone result. It is the first substantial evidence that Intel's pivot toward AI-capable server products and XPU accelerators is translating into actual revenue at meaningful scale. Intel announced strategic collaborations this quarter with Foxconn, Siemens, Hitachi, Echo Neurotechnologies, and Greenstone Biosciences, covering industry-specific AI and compute solutions built on Intel processors and purpose-built silicon.
The Foundry result is equally significant. Intel 18A-P entering risk production is the critical milestone the market has been waiting for since Lip-Bu Tan took over. Risk production means customer chips are being manufactured on the node at volumes sufficient to evaluate yield and performance at scale, which is the prerequisite for external customers to qualify the node for their own products. If yields continue improving, Intel's plan to become a credible TSMC alternative for leading-edge AI chips is no longer theoretical.
AMD's competitive announcement this week, EPYC Venice on TSMC's 2nm with 256 cores and Microsoft Azure as an early customer, underscores why Intel's 18A timeline matters. AMD's Diamond Rapids, the competing Xeon challenger, is not expected until mid-2027. That gives Intel roughly 12 months to establish the customer relationships and qualification pipelines that the Foundry business needs to demonstrate revenue conversion beyond internal products.
What Alphabet and Tesla's Selloff Was Really About
While Intel was producing its best quarter in 15 years, Alphabet and Tesla were delivering strong revenue numbers that the market punished anyway. Understanding the contrast explains what Thursday was actually about.
Alphabet beat on revenue at $119.8 billion, up 24%, and Google Cloud at $24.8 billion grew 82% with operating margin expanding to 35.6%. These are genuinely strong numbers. The stock fell 6.9% because Alphabet simultaneously raised full-year 2026 capex guidance to $195 to $205 billion, turned free cash flow negative for the first time in company history at minus $5.9 billion, and disclosed that Gemini 4 is being trained as a much larger model requiring more scale. The 82% cloud growth confirmed AI demand is real. The $205 billion capex number raised the question of when that demand converts to cash.
Tesla's Q2 revenue of $28.24 billion beat the $26.32 billion estimate. The stock fell 14.3% because adjusted EPS of $0.33 missed the $0.50 consensus by 34%, Musk described 2026 as a "massive capex year" on the call, and the CFO confirmed capex will grow for another two to three years. At a P/E above 349 times, the math of missing EPS while extending the capex runway is arithmetically punishing regardless of the underlying demand story.
The common thread is not that AI demand is broken. Google Cloud at 82%, Tesla's 1.48 million FSD subscriptions, and the strongest order backlog Tesla has seen since 2023 all point to demand that is intact. The thread is that the market has stopped extending credit on the promise that massive capital investment will eventually convert to proportional free cash flow. It wants to see the conversion begin, and Thursday's reports pushed that timeline further out.
Intel's result is structurally different. Intel is growing Data Center and AI revenue 59% on capex of $20 billion, not $205 billion. The capital intensity of Intel's AI revenue generation is orders of magnitude lower than the hyperscalers', and the operating leverage is beginning to show. That is the contrast Thursday buried under the Magnificent Seven headlines.
The Macro Context That Amplified Everything
The earnings results landed into a macro environment that was already positioned to punish any disappointment. Brent crude crossing $100 a barrel reintroduced the energy inflation channel the market thought it had escaped after June's ceasefire. The 10-year yield at 4.71% is the direct consequence, and it is the level at which valuation math for high-multiple growth stocks becomes structurally painful regardless of operating performance. Initial jobless claims hitting a 57-year low removed one of the remaining arguments for the Fed to hold at next week's meeting. Markets are now pricing above 33% probability of a July hike and above 78% for September.
In that environment, Alphabet's negative FCF and Tesla's EPS miss were not just earnings stories. They were read as confirmation that the highest-multiple components of the S&P 500 are simultaneously facing higher discount rates and extending their cash flow timelines. The Magnificent Seven have now generated essentially no collective return for 11 months, and with Microsoft, Meta, Amazon, and Apple all reporting next week, Thursday's framework is the one those results will be measured against.
Intel does not operate in that framework. It is not priced for perfection, it is not carrying $200 billion-plus capex ambitions, and it just demonstrated that AI infrastructure revenue can be generated at a capital intensity level the market has not yet properly valued. In a week when the Mag Seven lost $767 billion, that distinction matters.
Explore more exclusive insights at nextfin.ai.
