RockstarMarkets
All news
Markets · Narrative··Updated just now
Part of: AI Capex

NVDA Q2 Guidance at $91B Beats Consensus, Yet Stock Slips 2.5% After Hours

Despite a 140% YoY EPS surge and a $91B Q2 guide well above the $84-86B consensus, NVDA fell post-close as near-universal bullish positioning left little asymmetric upside, pressuring ^IXIC breadth across mega-cap tech.

R
Rocky · RockstarMarkets desk
Synthesised from 8 wires · 69 mentions in the last 24h
Sentiment
+30
Momentum
80
Mentions · 24h
69
Articles · 24h
88
Affected sectors
Related markets

Key facts

  • NVIDIA Q1 revenue $81.6B (+85% YoY), EPS $1.87 (+140% YoY), beat estimates
  • Q2 guidance $91B vs consensus $84-86B, driven by Blackwell deployment surge
  • Amazon adding 1M+ Blackwell and Rubin GPUs this year, $30-40B capex
  • Stock fell 2.5% after-hours despite beat; 20x forward PE, margin at cycle highs
  • $80B new buyback; H100 rental prices up 20% in 2026 despite three-gen-old chip

What's happening

NVIDIA delivered another blowout quarter that left little doubt about its dominance in AI infrastructure, yet the stock reaction exposed a critical tension: execution excellence is no longer enough to move the needle. The company reported revenue of $81.6 billion against analyst expectations of $74-79 billion, with data center revenue alone jumping to $75.2 billion, up 92 percent year-over-year. Earnings per share of $1.87 beat estimates by roughly 5 percent, marking a 140 percent year-over-year surge that underscores the scale of AI capex spending across hyperscalers.

The forward guidance offered the real surprise. NVIDIA guided Q2 revenue to $91 billion, far exceeding the consensus band of $84-86 billion and pointing to another quarter of explosive growth. The company attributed this surge to Blackwell demand, which is now entering deployment phase at major cloud operators. Amazon Web Services alone is adding more than 1 million Blackwell and Rubin GPUs this year, representing roughly $30-40 billion in chip purchases. The $80 billion buyback program and ongoing dividend commitments signal management confidence that cash generation will remain robust even as capex intensity rises across the semiconductor supply chain.

However, the after-hours decline of 2.5 percent highlights a deeper market dynamic. Sell-side consensus was uniformly bullish heading into the print, retail positioning was heavily long, and options markets were skewed bullish. A clean beat with a guide raise, while superficially ideal, may actually represent the worst asymmetric outcome for a stock already priced at roughly 20x forward earnings. Investors had already baked in strong guidance; the surprise was nowhere near the magnitude required to push a mega-cap index component higher. Some analysts flagged rising questions about whether hyperscaler capex can continue absorbing higher funding costs as bond yields hit their highest levels since 2007.

The broader implication cuts across semiconductors and AI infrastructure. NVIDIA remains the benchmark for investor appetite toward AI spending, and any moderation in hyperscaler capex plans would cascade through ARM, AMD, Broadcom, and smaller GPU-rental plays like CoreWeave and Lambda Labs. The fact that H100 rental prices are up roughly 20 percent in 2026 despite being three GPU generations old suggests supply-demand dynamics are still extremely tight. But the market's muted reaction to what many would call a perfect earnings print signals that the consensus may already be pricing in an extended cycle of strong growth without expecting further upside surprises.

What to watch next

  • 01Fed speakers and Treasury yields; 30Y at 5%+ may pressure capex budgets
  • 02Amazon, Microsoft, Google capex guidance in upcoming earnings
  • 03NVIDIA gross margins and guidance revisions over next two quarters
Mention velocity · last 24 hours
Coverage from these sources
Previously on this story

Related coverage

More about $NVDA

Topic hub
AI Capex: Who's Spending, Who's Earning, and What's at Risk

Tracking AI infrastructure capex — hyperscaler spend, data center buildouts, memory demand and the margin compression risk.