Intel Falls Twice as Hard as Its Chip Peers. That Gap Is the Real Signal

The market did Intel a small favor today by giving it company. The chipmaker fell about 5.6% to roughly $100, and its two largest rivals retreated too — Nvidia about 2.4%, AMD about 2.7% — after all three had just sprinted 11% higher over the prior five sessions. Read as a headline, this is profit taking: hot stocks, familiar wobble, buy-the-dip fodder.

Read as a gap, it is something else. Intel did not fall 3% like the others. It fell almost twice as hard. That gap between Intel and its peers is the signal, and it has been flashing all year.

The run that ate its own baseline

Intel is the strangest member of the AI chip trio because it arrived at this bull market from the cheap seats. Twelve months ago the stock traded near $24. It topped out around $142 this June — a roughly 500% run, by any measure parabolic — and now sits more than 30% below that peak. Today's dip is not the first crack; it is at least the second. Intel lost about 34% in July alone, its worst month since 2000, after that record rally, according to coverage of the company and the sector. The rebound into this week was momentum investors piling back in; today's pullback is that fuel running out of the tank.

The reason the drop matters is not that Intel is the headline victim. It is that Intel is the most exposed leg of a three-way trade that all three names are quoting off the same narrative: the AI spending boom is flowing past Nvidia into the "enablers" that make data centers run.

That framing made Intel, AMD, and even memory maker Micron rotate higher in tandem through the spring. In the second quarter alone, Micron climbed roughly 240%, Intel 216%, and AMD 186%, adding a combined $2 trillion in market value as investors widened their AI holdings beyond Nvidia. When three stocks quote off the same driver and fall together, that is not Intel pulling Nvidia down with it. That is a shared factor repricing the whole group — the AI-valuation wobble, not contagion. Two of the three can be right and the third can still be wrong about its own valuation.

The 6% versus the 3%

So why does Intel fall twice as hard? Because of what is holding the multiple up.

Look at the three on a price-to-earnings basis. AMD trades near $507 at a trailing price-to-earnings ratio above 125. Nvidia, the one name in the group with enormous delivered profit, trades at roughly 27 times trailing earnings. Intel is the outlier in the worst way: its trailing price-to-earnings ratio is negative, because on a GAAP basis it does not earn a profit at all. Its roughly $529 billion market capitalization is being justified by forward expectations and a story, not by current earnings.

And the recent "blowout" quarter shows how much of that is story. In the second quarter, Intel reported its strongest revenue growth in more than fifteen years — $16.1 billion, up 25% — with the Data Center and AI division jumping 59%. Then the GAAP line: a net loss of $2.16 per share, driven largely by a $12.5 billion non-cash charge tied to the CHIPS Act agreement with the U.S. government. Even stripping that out, the picture is a company growing fast but still unprofitable on the bottom line, with a foundry business that posted about $5.8 billion of revenue and a $2.1 billion operating loss.

That is the amplifier beneath the larger drop. The business Intel is betting its future on — the foundry — burns cash while capex climbs. Management raised full-year 2026 capital spending to about $20 billion and said 2027 would be higher. Net free cash flow over the trailing twelve months is about $2.8 billion against that spending. The market turned on those capex numbers the day they were raised: Intel beat Q2 estimates and still fell roughly 8% when it reported, as investors weighed the foundry losses and higher spending against a top line they had just applauded.

Here is the difference between Intel and its peers made concrete. Nvidia, at 25.8 times enterprise value to EBITDA, is expensive but earns the profit that backs the multiple — investors are paying up for cash that is actually being generated. AMD is far richer at about 86 times, but it is growing and profitable. Intel sits at about 45 times enterprise value to EBITDA with negative trailing GAAP earnings and thin cash flow. When the market's shared anxiety is "are AI valuations justified by delivered profit," the name with the least delivered profit per dollar of market cap wobbles first.

The firewalls that keep the chain short

None of this means Intel is the first domino in some collapsing AI tower. The chain only extends if the things holding the run up actually break, and right now the firewalls are real.

The most tangible one is the same one that started the rally: Intel is sold out of server CPU capacity through 2026, and average selling prices are reportedly rising 10% to 15% as AI workloads pull more traditional data center chips along with GPUs. That supply tightness is a legitimate earnings driver, not just a narrative, and it supports the idea that revenue growth can keep compounding even if the stock wobbles. The flip side is the threat: a stock that runs on tight supply trades on the supply staying tight. The chain stops the moment capacity fades, ASPs flatten, or AI capex doubt deepens — and Intel has no standing GAAP profit stream, the way Nvidia does, to hold its multiple on its own.

There are second and third props as well. The U.S. government took a near-10% stake in Intel last year through an $8.9 billion investment, and both Nvidia and SoftBank have put money in — a national-champion structure that gives the stock a floor of strategic buyers that ordinary chipmakers do not have. Reports of Apple exploring Intel chips for U.S. devices briefly spiked the stock 13% in May. Those are real backstops. They just do not show up on the income statement, which is exactly why they carry less weight on a day when the whole sector is repricing what the boom is actually worth.

What the gap is telling you

Strip the day's framing down to the part worth carrying. The word "profit taking" implies the story is intact and the dip is an entry. Sometimes that is true. But the market has now pulled the AI complex off its highs once, in July, on the same valuation anxiety that is flaring again — the SOX index fell into a bear market, down about 20% from its June record, as investors questioned how much of the AI capex boom would ever turn into returns, with China's Moonshot AI model feeding the doubt. Each pullback gets attributed to profit taking because it is the most comfortable word available, and each one is really a test of the same question: what is this worth if the profit is late?

For a retail investor, Intel's larger drop is the read. You do not need to own Intel to be exposed — at a $529 billion market cap it sits in the top tier of U.S. technology, so an index fund or a semiconductor ETF carries it and its volatility straight into your screen. The practical question the gap answers is which part of the AI-enabler trade is priced for perfection rather than priced for what it currently earns.

The chain does not need to break for the trade to have been overvalued at the top, and it does not need to break for Intel's 6% to have been the honest part of today's move. The first tripwire is the same one that fueled the rally: watch whether the server CPU supply stays sold out and whether selling prices keep rising. The amplifier is the foundry — capex spending and operating losses that keep cash flow thin. The decisive stop is the one test the market keeps running: if Intel can convert its record growth and tight supply into GAAP profit without the foundry swallowing it, the story earns its multiple and today becomes a footnote. If it cannot, the gap between a 6% drop and a 3% drop was never noise at all — it was the market telling you, months in advance, which name was carrying the least weight of its own.