This AI Giant Has Climbed 900% Over the Past Five Years. Now Wall Street Expects the Stock to Jump Another 40%.

Nvidia posted $81B in revenue in the recent quarter, with data center revenue up more than 90% to a record $75B and gross margin topping 70%, reinforcing its AI chip dominance. Wall Street is projecting roughly +40% upside over the next 12 months as Nvidia targets CPU expansion via its Vera Rubin platform, expecting $20B in stand-alone CPU revenue this year. Despite first-half stock gains of only 7% amid AI capex and potential in-house chip competition, the article frames valuation at 23x forward earnings as offering a renewed growth runway.

Analysis

Near term, this is mostly a quality-of-earnings rerating story for NVDA rather than a fresh fundamental step-change: the market still underestimates how much of its moat is software + systems integration, not just silicon. The bigger implication is that any CPU push is strategically defensive/offensive at once — it tightens lock-in around data-center architecture and makes switching costs higher, which should support gross margin durability more than headline unit revenue.

Second-order losers are INTC first and AMD second, but the real pressure point is data-center CPU share, not PC share. Even if NVDA takes only a modest slice of server CPUs, the signaling effect can force incumbents to spend more on roadmap execution and pricing, which can cap multiple expansion across the x86 group. The more important counterweight is that hyperscaler custom silicon at AMZN and META is still a latent leak in NVDA’s wallet share; that risk matters more over 6-18 months than over the next quarter.

Contrarianly, the consensus may be too eager to extrapolate the CPU initiative into a large incremental P&L driver. In agentic AI, the constraint is likely still memory bandwidth, networking, and power delivery — not CPU compute — so the $20B CPU target is better viewed as a bundling wedge than a stand-alone growth engine. If NVDA’s data-center growth decelerates materially or gross margin loses its 70% handle, the valuation support breaks quickly.

Key Decisions for Investors

• Long NVDA on 1-3 month pullbacks; treat the CPU roadmap and systems bundling as a margin-defense catalyst rather than a pure top-line story. Falsifier: next two quarters show data-center growth slowing sharply or gross margin slipping below the low-70s.
• Pair trade: long NVDA / short INTC over 6-12 months. The risk/reward favors NVDA because any AI-CPU traction hurts Intel’s highest-value franchise first, while NVDA’s downside is cushioned by GPU demand. Falsifier: Intel’s server roadmap or foundry commentary improves enough to stabilize share.
• Relative-value hedge: long NVDA, short AMD if the market starts pricing NVDA’s CPU move as a broad x86 share threat. AMD is better execution quality than INTC, so size this smaller; it works only if investor enthusiasm overstates the CPU opportunity.
• Watchlist, not a trade yet: AMZN and META. Their custom-chip efforts are the clearest medium-term ceiling on NVDA share, but the impact is more likely to show up as slower wallet share growth than outright displacement. Reassess if either company materially cuts external AI hardware spend.
• If already long semis, avoid adding broad ETF exposure here; favor NVDA concentration over SMH/SOXX until the market proves the CPU narrative is broad enough to lift the whole basket.