Memory Stocks Rout, Apple at $5T, SpaceX Slides: July 29

The AI chip sell-off that dominated late July shows no sign of letting up for semiconductor stocks, yet the broader S&P 500 is holding its ground. Meanwhile, two separate milestones are reshaping how investors think about the AI trade’s biggest beneficiaries.

Key points

• SK Hynix shares have been hit hard in what sources describe as a brutal summer for semiconductor stocks, with memory names bearing the brunt of the July sell-off.
• Micron and Sandisk are both under scrutiny as investors weigh which AI memory play, if either, is worth buying after the pullback.
• Apple briefly crossed a $5 trillion market cap on Tuesday, becoming only the second company ever to reach that level after Nvidia.
• SpaceX stock has slipped below key price levels following its mega-IPO, with at least one analyst flagging that negative sentiment may be obscuring long-term value.
• The Trump administration banned new Chinese humanoid robots and connected power inverters via the FCC, citing AI supply chain and cybersecurity risks.
• ChipAgents raised $60 million to use AI agents for chip design automation, bringing its total funding to $131 million.

Is the semiconductor sell-off a buying opportunity in memory?

July has been rough across the semiconductor sector, but memory chips have drawn particular attention. SK Hynix’s stock rout has prompted analysts to revisit the dynamics driving memory valuations, which ran hard on AI data center demand expectations. The question now is whether the sell-off reflects a genuine reassessment of AI infrastructure spending timelines, or whether it is a sentiment-driven overreaction.

On the U.S. side, a direct comparison between Micron and Sandisk has emerged as a key investor debate after both names pulled back sharply. Memory is central to AI workloads, particularly for high-bandwidth memory used in GPU clusters, but valuations had priced in aggressive growth. Notably, semiconductor stocks broadly are declining even as the S&P 500 remains resilient, suggesting the pain is sector-specific rather than a macro signal.

What does Apple at $5 trillion mean for the AI pecking order?

Apple’s brief crossing of the $5 trillion market cap threshold is a significant marker. It makes Apple only the second company ever to hit that level, after Nvidia. The fact that both companies in that club are deeply embedded in the AI hardware and device ecosystem underlines where the market sees durable value in this cycle. Nvidia got there first on data center GPU demand; Apple’s path runs through device-side AI integration and its services ecosystem.

For investors tracking the AI investment cycle, the contrast with the semiconductor sell-off is striking. Broad chip names are under pressure, yet the two largest companies in the world by market cap are both AI infrastructure and AI device plays. The divergence suggests the market is becoming more selective, rewarding companies with clear monetization paths while punishing those exposed to capex cycle uncertainty.

SpaceX post-IPO slide and the Chinese robot ban

SpaceX continues to trade below its IPO price, with sources noting the stock has been questioned at levels below $120 and one analyst warning about the risks if it falls below $100. Post-IPO pressure is common, but the size of the drawdown has prompted debate about whether the initial valuation priced in too much optimism. SpaceX also faces a separate legal distraction: xAI and SpaceX are contesting claims in a noise lawsuit tied to a power plant, calling the allegations flawed.

On the policy front, the FCC’s ban on new Chinese humanoid robots and connected power inverters has direct implications for AI hardware investors. The measures apply to models not yet released and took effect immediately. Unitree, which holds roughly a fifth of global humanoid robot market share according to Counterpoint Research, is expected to be among the most affected. Unitree had recently partnered with Nvidia to use Blackwell chips in its robots. Nvidia has stated that data from those robots stays in the U.S. The ban also covers connected power inverters, a category critical to data center energy infrastructure, where reliable non-Chinese sourcing is now a strategic priority.

A funding bright spot in chip design

Away from the volatility, ChipAgents closed a $60 million funding round to expand its AI-agent-driven approach to semiconductor design and verification. The Nvidia partner has now raised $131 million in total. Chip design automation is a long-cycle bet: if AI agents can meaningfully compress verification timelines, the downstream beneficiaries would be every company waiting on custom silicon. It is early-stage, but the investor appetite for the category remains intact even as public semiconductor stocks sell off.

Nothing on this site constitutes investment advice. Data and analysis are based solely on the sources cited.