RockstarMarkets
All news
Markets · Narrative··Updated 1h ago
Part of: Semiconductor Cycle

AI Capex Boom Keeps Memory Chips Expensive; NVDA, SMCI Defy Valuation Logic With Record Pricing

Memory chip makers are commanding record valuations despite scorching price rallies as insatiable AI demand sustains capex spending. NVIDIA has climbed to $5.5 trillion in market cap on AI infrastructure tailwinds, while Super Micro and chipmakers defy typical mean-reversion logic, with high forward multiples justified by AI buildout momentum that shows no signs of slowing.

R
Rocky · RockstarMarkets desk
Synthesised from 8 wires · 46 mentions in the last 24h
Sentiment
+70
Momentum
80
Mentions · 24h
46
Articles · 24h
22
Affected sectors
Related markets

Key facts

  • NVIDIA reached $5.5 trillion market cap on sustained AI infrastructure demand
  • Meta signed $21 billion deal with CoreWeave for long-term AI inference capacity
  • Memory chip makers showing record valuations despite high price rallies; earnings growth justifies multiples
  • Broadcom, AMD supply constraints flagged as potential capex bottlenecks
  • SMCI, NVDA trading at elevated multiples; AI infrastructure capex expected through 2027+

What's happening

The artificial intelligence infrastructure build-out is defying traditional valuation discipline and sustaining memory chip rallies at levels that would normally trigger profit-taking. Bloomberg analysts note that some of the semiconductor industry's top stocks have become cheaper to own in traditional metrics (lower forward P/E ratios) even as their share prices reach nosebleed levels, a paradox explained by persistently high earnings growth driven by AI capex. NVIDIA alone has reached a $5.5 trillion market cap, with CEO Jensen Huang spotted walking Beijing's streets during the Trump-Xi summit, symbolizing the company's outsized influence over tech narratives.

Meta has just signed a $21 billion agreement with CoreWeave for long-term AI inference capacity, a structural shift beyond just model training. This multi-year capex commitment from one of the world's largest tech players is anchoring expectations that AI infrastructure demand will persist well into 2027 and beyond. Memory chip availability is becoming a bottleneck; Goldman Sachs research notes that Broadcom chips may present a constraint in some AI deployments, while AMD supply in switches is expected to scale meaningfully over the next 2-3 years. These supply dynamics are supporting pricing power and valuation premiums for incumbents like NVDA and SMCI.

The narrative also touches on geopolitical tailwinds: reports that the US approved Chinese companies to buy Nvidia chips (later disputed) briefly sent NVDA "parabolic" in social chatter, highlighting how trade policy uncertainty and China openings still move the tape. Investors are crowded into these names, and consensus is now betting on sustained AI infrastructure spending through at least 2027, with little room for disappointment.

The key risk is capex satiation: if major cloud providers announce capacity sufficiency or if spending growth slows materially, memory stocks could see a sharp repricing. Additionally, any further tightening of China export restrictions would pressure NVDA's revenue guidance and shake confidence in the durability of the rally.

What to watch next

  • 01Meta, Google, Amazon quarterly capex guidance: Q2 earnings season (late April-May)
  • 02NVIDIA earnings and China export policy developments: Q2 2026 report (late May)
  • 03Broadcom, AMD capacity utilization updates: Q2 results and management commentary
Mention velocity · last 24 hours
Coverage from these sources
Previously on this story

Related coverage

More about $NVDA

Topic hub
Semiconductor Cycle: AI Capex, Memory and the SOX Trade

Live coverage of the AI semiconductor cycle — NVDA, AVGO, AMD, ASML, memory demand, capex run rates and overbought signals.