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

AI Semiconductor Rally Continues; Memory Stocks Most Expensive Since 2021 Despite Scorching Performance

Memory and AI semiconductor stocks have reached valuation peaks despite surging share prices, with Bloomberg noting they are becoming cheaper on relative metrics even as absolute prices soar. Nvidia breakout past $5.5 trillion market cap reflects both AI infrastructure conviction and crowded positioning, raising questions about peak capex risk.

R
Rocky · RockstarMarkets desk
Synthesised from 8 wires · 44 mentions in the last 24h
Sentiment
+60
Momentum
80
Mentions · 24h
44
Articles · 24h
23
Affected sectors
Related markets

Key facts

  • Nvidia market cap exceeded $5.5 trillion; stock rallied $22 intraday on AI momentum
  • Memory stocks most expensive since 2021 on relative P/E despite price surges
  • Meta committed $21B to CoreWeave for inference capacity; critics question ROI
  • Broadcom flagged as potential supply constraint for AI infrastructure expansion
  • "Peak capex" debate intensifying as sell-side strategists warn of deceleration risk

What's happening

The AI semiconductor rally has entered a phase where price performance is outpacing fundamental valuation expansion. Memory chip stocks like Micron (MU) and broader semiconductor names are now trading at elevated multiples relative to historical norms, yet their stock prices continue to push higher. This divergence, rising stock prices amid compressed valuations, typically signals crowded crowded trades where momentum and passive flows are driving prices more than earnings growth. Nvidia's breakthrough past $5.5 trillion market cap is both a testament to AI infrastructure demand and a red flag for conviction traders.

The narrative around AI capex remains contested. Sell-side strategists are warning of "peak capex" risk, suggesting that the historic levels of spending on data centers, GPUs, and training infrastructure may not sustain forever. Meta's $21 billion commitment to CoreWeave for inference capacity is being scrutinized: is this expansion or consolidation of already-announced budgets? Cisco's recent earnings call hinted at sustained networking demand for AI buildout (switches, optics, routing), but the commentary lacked the exuberance of earlier quarters. Broadcom (AVGO) could become a constraint if chipmakers cannot source critical passive components.

Institutional positioning is mixed. Some passive index trackers are overweighting AI mega-caps simply due to market-cap indexing; Palantir (PLTR) and Super Micro (SMCI) have attracted retail attention but remain controversial for governance and accounting. Raymond James analysts flagged AMD and Arista (ANET) as potential next-gen beneficiaries if customers diversify away from Nvidia monoculture, but adoption timelines remain uncertain.

The bearish case hinges on capex deceleration or disappointing ROI. If major cloud providers (AWS, Azure, Google Cloud) report underwhelming returns on AI infrastructure spending, capex budgets could contract sharply, pressuring Nvidia, AMD, and supporting foundries. Nvidia's China export restrictions (if reversed) could spark margin competition as Beijing-based chipmakers gain parity in supply chains.

What to watch next

  • 01Nvidia earnings call June 2026: capex guidance and customer demand signals
  • 02Meta, Amazon, Google cloud earnings in Q2 2026: AI infrastructure ROI disclosure
  • 03AMD's share-gain strategy: can EPYC processors dent Nvidia's GPU monopoly?
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.