More than $1 trillion in market capitalization vanished from the world’s most valuable semiconductor and memory companies in a single week of late July 2026, as investor anxiety over AI infrastructure spending finally caught up with a sector that had seemed immune to gravity. Nvidia alone absorbed a $238 billion rout since the prior Friday’s close, while memory giants SK Hynix, Samsung Electronics, and Micron lost $176 billion, $173 billion, and $113 billion respectively, according to figures reported by CNBC.
What Triggered the Selloff
The proximate cause was a convergence of worries: stretched valuations after nearly three years of uninterrupted AI-driven gains, renewed scrutiny of circular financing arrangements between chipmakers and AI labs, and jitters ahead of a dense earnings calendar featuring Alphabet, Tesla, Intel, Meta, Amazon, and Microsoft. The Nasdaq-100 slid into correction territory as the selloff spread beyond pure-play chip names into the broader memory and equipment supply chain, according to NBC News reporting on the global rout.
Memory Makers Bear the Brunt
Notably, the damage wasn’t confined to logic chipmakers like Nvidia. Memory suppliers SK Hynix and Samsung, which have ridden a historic high-bandwidth memory (HBM) boom tied directly to AI accelerator demand, saw some of the sharpest percentage declines. That’s significant because HBM has been one of the tightest-supplied components in the entire AI hardware stack for the past two years, with lead times stretching well into 2027 for some customers. A sudden repricing of that segment suggests investors are questioning not the demand itself, but the multiples they had been willing to pay for it.
How We Got Here
The chip sector’s run-up traces back to the ChatGPT moment in late 2022, but the current phase of investor mania accelerated through 2025 as hyperscalers committed hundreds of billions of dollars to data center buildouts. Valuations across the space, from foundries to memory to packaging equipment, climbed to levels that assumed years of uninterrupted double-digit growth. Any wobble in that assumption, whether from a single disappointing earnings report or a broader macro shift, was always likely to trigger an outsized correction given how much money had piled into the trade.
Bulls See a Buying Opportunity, Bears See a Reckoning
Some strategists frame the pullback as a healthy, overdue reset rather than the start of a structural downturn, noting that underlying demand signals, order backlogs, and capital spending guidance from hyperscalers have not meaningfully changed. Others argue the selloff is the first real crack in the AI capex thesis, pointing to earlier warnings about circular deals, in which chipmakers invest in AI startups that then use the proceeds to buy chips, potentially inflating reported demand. Skeptics also note that a trillion-dollar swing in a single week reflects how thin the margin for disappointment has become across the sector.
The Earnings Gauntlet Ahead
The timing amplifies the stakes: this rout hit just as Alphabet and Tesla were due to report, with Intel following shortly after, and Meta, Amazon, and Microsoft close behind. Each of those companies is a major buyer of AI infrastructure, and any commentary suggesting a pullback or deceleration in capital spending plans could deepen the selloff further. Conversely, reaffirmed or raised spending guidance could help stabilize sentiment across the semiconductor complex.
What It Means Next
The rout is a reminder that even the most dominant growth story in modern markets isn’t immune to repricing when expectations run too far ahead of near-term fundamentals. For chipmakers, the coming weeks of earnings calls will be a referendum on whether AI infrastructure demand remains as durable as the market has priced in, or whether 2026 marks the year the AI capex supercycle starts to show its first real strain. Either way, expect volatility across chip and memory stocks to stay elevated well into the fall.
Beyond institutional trading desks, the selloff has been felt acutely by retail investors who piled into semiconductor exchange-traded funds and individual chip stocks during the sector’s multi-year run. Several popular leveraged chip ETFs recorded some of their steepest single-week declines since their inception, a reminder that products designed to amplify gains during a bull run can just as easily amplify losses when sentiment turns. Financial advisors have reported a surge in client inquiries about rebalancing portfolios that had become unusually concentrated in a small handful of AI-adjacent names.
The episode also reignited a long-simmering debate among market strategists about concentration risk in major indices. With chip and AI-infrastructure names representing an outsized share of both the S&P 500 and Nasdaq-100’s market capitalization, a selloff of this magnitude in a single sector can distort the performance of index funds far beyond what a typical investor might expect from a diversified holding, underscoring how intertwined the broader market has become with the fortunes of a handful of semiconductor giants.