Nvidia stock price target reiterated at $250 by Goldman Sachs

Goldman Sachs reiterated a Buy rating on Nvidia with a $250 price target, citing upside to its $1 trillion datacenter guidance, potential agentic AI-driven demand, and a beat-and-raise quarter. The firm raised estimates by about 12% on average, with CY26/27 estimates now 14%/34% above Street, though it noted stock outperformance may be limited if investors demand clearer proof of improved hyperscaler profitability and broader enterprise adoption.

Analysis

The setup is less about a single-quarter beat and more about whether the AI capex cycle can transition from hype-driven to infrastructure-validated. The key second-order signal is that multiple ecosystem players are still coordinating around networking, inference, and cluster efficiency, which argues the bottleneck is shifting from “can demand exist?” to “can deployments scale economically?” That usually supports the whole compute stack, but it also increases scrutiny on whether NVDA’s premium can survive if hyperscaler ROI metrics do not inflect over the next 1-2 quarters.

The biggest competitive nuance is that the marginal beneficiaries are no longer only GPU vendors. If agentic AI meaningfully raises server CPU attach rates, Intel and AMD can participate even without displacing Nvidia, while Broadcom benefits if cluster interconnect and custom silicon spending continues to compound. In other words, the market may be underestimating how much of the next leg is a breadth trade across infrastructure rather than a single-name dominance trade.

Near term, the risk is not demand collapse but disappointment versus a very elevated narrative bar. If management commentary implies datacenter upside is timing-shifted rather than demand-constrained, the stock can de-rate despite solid fundamentals. Over a 3-6 month horizon, the decisive catalyst is evidence from hyperscaler margins: if capex remains high while profitability improves, multiples expand; if margins stall, the market will start treating AI spend as financing risk rather than growth optionality.

The contrarian view is that NVDA may be cheap on trailing metrics but not on the forward expectations embedded in consensus and positioning. The broader AI basket has already been bid on a narrow leadership narrative, so the cleaner risk/reward may be in lagging adjacencies with less valuation fragility and earlier estimate revision cycles. That favors selectively owning the picks-and-shovels beneficiaries while fading the idea that every good AI datapoint automatically translates into further NVDA multiple expansion.

Key Decisions for Investors

• Maintain a tactical long NVDA only into the print/next catalyst window; use upside-defined call spreads rather than spot equity to express a beat-and-raise view, since the stock likely needs a meaningful guide-up to outperform.
• Pair trade: long AVGO / short NVDA for 1-3 months if you expect AI infrastructure breadth to matter more than single-name multiple expansion; AVGO has cleaner exposure to interconnect/custom silicon and less earnings sensitivity to headline valuation swings.
• Add AMD on pullbacks as a secondary beneficiary of broader AI compute demand; use it as a higher-beta expression of continued datacenter spend with better upside if server CPU attach and inference breadth improve over the next 2-4 quarters.
• Consider a modest long INTC as a contrarian catch-up trade if management commentary across the ecosystem confirms rising CPU importance in AI servers; risk/reward improves only if the market starts pricing AI non-GPU content growth.
• Avoid chasing the name outright ahead of earnings if positioning is crowded; instead wait for post-print confirmation of hyperscaler margin inflection, then add on a 5-8% pullback or on upward estimate revisions.