#IntelQ2RevenueSurges25%

Intel Didn't Just Beat Earnings It Rewrote the Narrative

Twelve months ago, Intel was the punchline of every semiconductor conversation. Missed mobile. Late to AI. Foundry hemorrhaging cash. The stock was left for dead. Then Lip-Bu Tan walks in, guts the bureaucracy, and starts building something different.

$16.1 billion in Q2 revenue. Not a gentle beat a $1.7 billion demolition of the $14.4 billion consensus. That's the fastest top-line growth Intel has posted since Q3 2011, when the world was still buying desktops and the iPhone was four years old. Twenty-five percent year-over-year. From a company that was shrinking quarters ago.

Non-GAAP EPS of $0.42 against a $0.22 Street estimate. Not a rounding error nearly double. Non-GAAP net income swung to $2.2 billion from a $441 million loss a year ago. Operating margin vaulted from negative territory to 17.2%. Gross margin recovered to 41.8%, up twelve full percentage points from the year-ago quarter's 29.7%.

Data Center & AI hit $6.3 billion, up 59% year-over-year. That's more than double the company's overall growth rate. Xeon 6 is being described as one of the fastest-ramping products in Intel's history. For a division that was supposed to be eating Nvidia's dust, that's a statement. DCAI is now throwing off a 40% operating margin. The CPU is not dead agentic AI and inference workloads are pulling demand back toward the architecture Intel has always dominated.

Intel Foundry hit $5.8 billion, up 31%. That's nearly double the 16% growth seen in Q1. The 18A node is running 25% above target, with output up over 50% quarter-over-quarter. Fortinet just signed on as the first named external foundry customer under Tan's tenure. The foundry still lost $2.1 billion this quarter but that loss narrowed by more than a billion dollars year-over-year. The trajectory matters more than the absolute number right now.

Client Computing & Physical AI delivered $8.9 billion, up 13%. AI PCs are holding the line, even as memory shortages bite at supply.

Q3 guidance: $15.8–16.8 billion vs. $15.06 billion consensus. Non-GAAP EPS of $0.38 vs. $0.27 expected. The company is not just having a good quarter it's projecting forward momentum.

Three reasons. First, the GAAP numbers tell a different story: a $2.16 per-share loss, driven by a $12.5 billion non-cash charge tied to shares owed to the U.S. government. That's accounting, not operations but it spooks retail investors who don't read footnotes. Second, Intel is raising capex to over $20 billion for 2026, with 2027 "significantly above" that. That's a massive cash burn signal, and the stock is already up 170% year-to-date. The re-rating is largely priced in. Third, analysts are staying cautious. Bernstein kept market-perform. One firm even raised its price target to $80 while maintaining a Sell rating. The 14A node the real foundry catalyst still hasn't landed a major external customer. The turnaround is real, but the valuation leaves little room for error.

Lip-Bu Tan's framing is worth noting: "AI-driven compute continues to strengthen, and to support expected growth this year and next across products and foundry, we are meaningfully increasing our investments in equipment, clean room space, and substrates." He's also said memory shortages will persist until at least 2028. This is not a CEO cashing in a victory lap it's a CEO loading up for a multi-year war.

The bottom line: Intel's operating turnaround is no longer theoretical. The numbers are real. The margin expansion is real. DCAI is growing at rates that warrant serious attention. But the stock is pricing in a lot of hope, and the foundry's path to profitability is still years away. The smart play here is to separate the business momentum from the valuation narrative they're telling two very different stories right now.