Nvidia stock jumps as earnings top Wall Street estimates

Nvidia reported Q1 revenue of $81.6B versus $79.15B expected and adjusted EPS of $1.87 versus $1.77, while Q2 revenue guidance of $91.0B topped the $87.2B consensus. The company also authorized an additional $80B share buyback and raised its quarterly dividend from $0.01 to $0.25 per share. Offset by continued China restrictions and uncertainty around Middle East supply-chain risks, the print remains supportive for the AI trade but leaves export-control risk unresolved.

Analysis

The most important read-through is not the beat itself but the quality of demand signal: guidance upside while China is effectively zero means the core AI buildout is still absorbing capacity faster than the market can discount. That shifts the debate from near-term order cyclicality to durability of capex, and it keeps NVDA positioned as the toll collector on a multi-year infrastructure spend rather than just a beneficiary of hype. The buyback/dividend step also tells you management sees limited near-term friction in converting earnings to cash, which supports the stock on any post-print digestion.

Second-order, the competitive threat is becoming more architectural than product-level. The real risk is that hyperscalers use the next 6-12 months to migrate a growing share of inference and lower-end training to in-house silicon, which compresses NVDA's mix and bargaining power even if overall AI spend stays elevated. That dynamic is more likely to cap upside in AMZN than to materially impair NVDA in the next two quarters, because the latter still owns the performance frontier while the former is trying to prove compute monetization on a cost basis.

China remains the largest unresolved optionality and the market is underpricing how binary it is. If licenses remain unmonetized, the upside case is already being driven entirely by ex-China demand; if shipments reopen, it is incremental rather than thesis-changing. The geopolitical filing also matters because it creates a tail-risk discount that can widen quickly if Middle East supply-chain disruptions or export-control escalation start touching lead times; this is a months-not-days issue, but it is a credible source of multiple compression if sentiment in AI overheats.

Contrarian view: the consensus is still treating all AI names as one trade, but NVDA’s print argues for dispersion. The beneficiaries of rising AI spend are increasingly the model owners and infra enablers with independent demand, while the hardware ecosystem may face margin pressure as customers diversify suppliers and internalize more chips. That makes the current setup better for relative value than outright beta chasing.

Key Decisions for Investors

• Stay long NVDA on any 2-3% post-earnings pullback over the next 1-2 sessions; use the print-confirmed demand upgrade as a tactical add, with a 5-7% downside stop if guidance optimism fades on follow-through commentary.
• Pair trade: long NVDA / short AMZN over the next 1-3 months. NVDA has cleaner ex-China monetization and superior pricing power; AMZN faces the higher risk that in-house silicon capex displaces third-party GPU demand, making this a good relative-value expression.
• Initiate a small short or put spread in AMD into strength over the next 2-6 weeks. The stock can benefit from share gains, but this quarter reinforces that the frontier remains NVDA-led, and partnership headlines may not translate into near-term shipment leverage.
• Buy NVDA upside calls only if implied vol compresses after the open, targeting a 3-6 month horizon. The thesis is not a single-print pop but continued AI capex durability; use call spreads to cap premium and define risk.
• Avoid overexposure to AMZN until there is proof Trainium monetization is additive rather than substitutive; if hyperscaler silicon adoption accelerates, the risk/reward favors owning the chip leader over the customer-turned-competitor.