Nvidia reported its FY2026 fourth-quarter earnings on February 25, crushing Wall Street estimates across all key metrics. Revenue reached $68.1 billion, up 73% year-over-year and 20% sequentially, setting a new all-time high. Non-GAAP EPS came in at $1.62, above the $1.53 consensus estimate. Following the release, Morgan Stanley raised its price target from $250 to $260, while RBC Capital Markets increased its target from $240 to $250, both maintaining an Overweight rating.
Data Center Revenue Hits $62.3B, Blackwell Dominates Inference
The data center segment remains the primary growth engine, generating $62.3 billion in quarterly revenue, up 75% year-over-year and 22% quarter-over-quarter, representing over 91% of total revenue. Full-year data center revenue reached $193.7 billion, up 68% year-over-year, signaling no slowdown in AI compute demand from global hyperscalers. CEO Jensen Huang stated on the call: "Grace Blackwell with NVLink is the king of inference today, reducing the cost per token by an order of magnitude." He also previewed the next-generation Vera Rubin platform, which will further cut inference costs to one-tenth of Blackwell's, cementing Nvidia's dominance in AI infrastructure.
Q1 Outlook of $78B Far Exceeds Estimates
Wall Street was most impressed by the guidance: Nvidia expects FY2027 first-quarter revenue of $78 billion (±2%), well above the analyst consensus of $72.6 billion, suggesting that the AI spending cycle is accelerating, not peaking. Non-GAAP gross margin is forecasted at 75.0% (±50 bps), with operating expenses around $7.5 billion.