RockstarMarkets
All news
Markets · Narrative··Updated 1h ago
Part of: AI Capex

NVDA Q2 Guidance at $91B Signals AI Capex Cycle, But AH Selloff Raises the Bar

Nvidia's Q1 beat of $81.6B revenue and Q2 guide of $91B cleared every consensus hurdle, yet shares fell 2.5% after hours as peak-positioning eroded asymmetric upside. H100 rental prices up 20% despite three-generation obsolescence flags sustained scarcity, pressuring hyperscaler margins and widening the capex sustainab

R
Rocky · RockstarMarkets desk
Synthesised from 8 wires · 68 mentions in the last 24h
Sentiment
+40
Momentum
85
Mentions · 24h
68
Articles · 24h
83
Affected sectors
Related markets

Key facts

  • NVDA Q1 revenue $81.6B (+85% YoY), data center $75.2B (+92% YoY)
  • Q2 guidance $91B vs. consensus $84-86B; EPS $1.87 vs. $1.78 estimate
  • H100 GPU rental prices up ~20% in 2026 despite being 3 generations old
  • NVDA stock fell 2.5% AH despite beat and raise; options skewed bullish
  • $80B buyback and dividend expansion; analyst calls question capex sustainability

What's happening

Nvidia delivered its strongest earnings yet on May 21, posting Q1 revenue of $81.6B against expectations of $79.2B and EPS of $1.87 versus $1.78 forecast. Data center revenue surged to $75.2B, up 92% year-over-year, underscoring relentless demand for Blackwell GPUs from hyperscalers building out AI infrastructure. The company issued Q2 guidance of $91B, significantly above consensus expectations of $84-86B, and announced an $80B share buyback program with an expanded dividend. Jensen Huang's commentary on Blackwell ramp-up and upcoming Rubin platform validated the AI capex thesis, even as gross margins held steady and power, cooling, and optics emerged as emerging constraints on data center deployments.

Yet the market's response was muted. NVDA shares fell 2.5% in after-hours trading despite the beat and raise, a tell that investor positioning had already priced in perfection. Options markets showed skewed bullish OI and heavy retail long exposure ahead of the release, meaning the asymmetric risk cut both ways. Sell-side analysts remained uniformly bullish, but on the call Huang fielded questions about the viability of sustaining 90%+ quarterly revenue growth rates amid rising capex costs for hyperscalers. Bank of America and Goldman Sachs research highlighted the bar being set so high that execution risk now exceeds upside surprise potential, particularly if Amazon, Microsoft, or Google recalibrate AI spend in response to macro pressure or rising bond yields.

The narrative has shifted from AI adoption to AI capex saturation risk. Large institutional holders, including BlackRock, began repositioning allocations across the semiconductor supply chain. H100 GPU rental prices rose ~20% in 2026 despite the chips being three generations old, signaling sustained scarcity but also inflating the user economics for downstream data center operators. Energy importers and utilities face margin compression from elevated power demand tied to AI rack density, while the downstream beneficiaries, Broadcom, AMD, Marvell, ARM, may outperform on relative valuation if NVDA's growth rate normalizes.

Skeptics point to the mounting debt and capex burden now shouldered by hyperscalers themselves. Goldman Sachs warned that if cost-of-capital trends persist or if enterprise-level AI ROI remains elusive, hyperscaler capex could stall, a scenario that would hollow out Nvidia's forward guide and trigger sector-wide rotation. The fact that NVDA held near $223 despite the guide raise suggests the market is pricing in margin compression from rising utility and financing costs ahead.

What to watch next

  • 01Hyperscaler capex guidance: AMZN, MSFT, GOOGL earnings for Q2 AI spend trends
  • 02Fed policy and bond yields: rising rates pressure capex ROI math
  • 03NVDA margin trends: gross margin defense amid power and cooling cost escalation
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.