Major Hyperscalers Just Reported Earnings. Nvidia Was The Winner

Alphabet, Microsoft, and Amazon are increasing AI-related capex, with Microsoft guiding to $190 billion in capex for 2026 and Amazon having spent $43.2 billion year-to-date against a $200 billion plan for 2024. The article argues this sustained spending supports Nvidia’s chip demand, even as Amazon expands its own Trainium AI chips, because the cloud leaders still plan to buy substantial quantities of Nvidia hardware. Overall, the piece is constructive for Nvidia’s long-term revenue and earnings outlook, but it is primarily opinion-driven rather than a new company-specific catalyst.

Analysis

The key second-order effect is that hyperscaler capex is becoming less cyclical and more competitive-defensive: once one cloud platform leans harder into AI infrastructure, the others cannot easily slow without risking share loss in enterprise workloads. That creates a self-reinforcing spending ladder that should keep Nvidia’s order book firmer for longer than bearish semiconductor investors expect, especially because deployment decisions are now tied to customer retention, not just ROI math.

The market is still underestimating mix shift risk inside the cloud stack. Amazon’s internal silicon can compress Nvidia’s share of wallet over time, but that is more likely to change pricing power than volume in the near term because the most latency-sensitive and model-training workloads will still route to best-in-class GPUs. The real beneficiary set is broader than NVDA alone: networking, power, and advanced packaging suppliers can capture incremental spend as data-center density rises and power constraints become the binding bottleneck.

The contrarian risk is timing: the demand narrative is strong for the next several quarters, but the setup becomes more fragile in 2026-2027 when capex ramps collide with depreciation, energy costs, and investor pressure on cloud margins. If hyperscalers begin optimizing for capital efficiency rather than capacity buildout, Nvidia’s growth rate can decelerate sharply even if units remain healthy. That makes the stock attractive, but also vulnerable to a valuation reset if guidance implies a slower cadence than the market has already priced.

From a positioning standpoint, this reads as a bullish continuation rather than a fresh re-rating catalyst. The best trade is not simply long NVDA outright; it is long NVDA versus the weaker monetization names that are spending heavily but lack the same pricing power, and against any assumption that internal chips will quickly displace GPU demand. The market is probably overdiscussing substitution and underweighting the fact that AI infrastructure spend is expanding faster than single-vendor replacement curves.

Key Decisions for Investors

• Maintain/add long NVDA on any post-rally consolidation over the next 2-6 weeks; target a 6-12 month hold, with upside from sustained hyperscaler capex visibility and downside mainly from any 2026 guidance miss.
• Pair trade: long NVDA / short INTC over 3-9 months. Rationale: NVDA captures the premium AI training spend, while INTC remains a cleaner beneficiary of legacy compute substitution but with weaker AI monetization; skew favors NVDA if capex stays expansionary.
• Express the ‘picks-and-shovels’ follow-through via long ANET or CLS versus short a broad semis basket for 3-6 months; if data-center density keeps rising, networking and high-speed interconnect should see second-order demand even if GPU share shifts modestly.
• Buy medium-dated NVDA call spreads into any dip over the next 1-2 weeks rather than outright calls; this preserves upside through the next hyperscaler-guidance cycle while limiting premium bleed if the market digests the already-bullish capex narrative.
• Watch for any confirmation that 2026-2027 capex plans are accelerating; if so, add to NVDA and consider reducing exposure only if hyperscaler margin commentary turns defensive, which would be the earliest signal that spending is peaking.