Burry Wins Chip Bet, ON Semi Jumps, TSMC N3 Tightens

This update is a roundup of same-day reporting from the linked sources below, with editorial context from the CPJ Stock Desk.

A volatile week for chip stocks produced at least one clear winner: Michael Burry, who bet against the sector just before a brutal July selloff. Meanwhile, ON Semiconductor offered a rare bright earnings signal, and a TSMC capacity crunch is drawing fresh attention to Intel’s foundry ambitions.

Key points

Michael Burry disclosed a short position against the iShares Semiconductor ETF (SOXX), which fell 21% in July, appearing to validate the trade.
ON Semiconductor shares jumped 7.6% in pre-market trading after Q2 2026 results and Q3 guidance both beat Wall Street forecasts.
Micron (MU) is still up roughly 190% in 2026 but has dropped more than 25% from its late-June all-time high, including an approximately 9% single-day plunge on July 28.
TSMC’s N3 node capacity may remain constrained for years as Nvidia and Google ramp AI chip production, opening a potential opening for Intel’s 18A and 14A process nodes.
Nvidia is leading a $74 million extended Series B round for Indian AI startup Sarvam AI, committing roughly $25 million of its own capital.

Did Burry nail the chip short?

The “Big Short” investor publicly stated he had bet against SOXX, the iShares Semiconductor ETF, and the fund subsequently dropped 21% through July. Whether the position is still open or has been closed is not confirmed in available sources, so the final profit, if any, remains unknown. Still, the timing aligns closely enough that the trade is drawing significant attention.

The broader context matters here. Micron’s roughly 9% single-day drop on July 28 was part of a wider chip stock selloff that hit the sector hard. Micron remains one of the year’s strongest performers at approximately 190% gains through early August, but the 25%-plus pullback from its June peak illustrates just how sharp reversals can be in AI-adjacent semiconductor names. Investors comparing Micron to ASML face a genuine risk-reward question: one stock has captured enormous upside tied directly to AI memory demand, the other offers more stable exposure to the equipment layer underpinning the entire industry.

What does the TSMC N3 crunch mean for investors?

The N3 node at TSMC is under pressure. Nvidia and Google are both ramping AI chip production on that process, and TSMC has warned that wafer capacity could stay tight for years. That scarcity has two implications. First, it gives TSMC meaningful pricing power, which is constructive for TSM shareholders. Second, it creates an opening for Intel’s foundry business, which is pitching its 18A and 14A nodes as an alternative for customers who cannot get enough TSMC capacity.

Whether Intel can realistically absorb overflow demand is a separate and debated question. Intel’s foundry ambitions have faced years of execution challenges, and winning large AI chip customers requires more than available capacity. It requires yield, trust, and competitive performance specs. The N3 squeeze at TSMC does, however, give Intel a more credible sales conversation than it has had in some time.

ON Semiconductor breaks from the pack

Not every chip name is struggling. ON Semiconductor reported Q2 2026 results that cleared Wall Street expectations and paired them with a stronger-than-anticipated Q3 outlook, sending shares up 7.6% in pre-market trading. ON Semi’s business is more weighted toward industrial and automotive end markets than pure-play AI silicon, which may partly explain why it avoided the worst of the sector’s July turbulence and is now rewarding investors with a beat-and-raise quarter.

Nvidia bets on Indian AI infrastructure

Away from the public markets, Nvidia is leading a $74 million extended Series B round for Sarvam AI, a Bengaluru-based startup that recently crossed the unicorn threshold. Nvidia’s own contribution is approximately $25 million, according to regulatory filings cited by Entrackr. Glade Brook Capital and Gaja Capital are also participating. Sarvam AI is working on large foundation models, and Nvidia’s involvement underscores the company’s strategy of investing directly in AI model developers who will, in turn, consume its hardware. It is a customer-development play as much as a financial one.

This is not investment advice. All figures are sourced from the articles cited and have not been independently verified.