Nvidia Revenue Hits New Record High, Jensen Huang Says AI Agent Inflection Point Has Arrived

On February 25th after the U.S. stock market closed, NVIDIA announced its financial results for the fourth quarter of fiscal year 2026 and the full year ending January 25, 2026.

Data shows that NVIDIA’s Q4 revenue hit a record high of $68.1 billion, up 20% quarter-over-quarter and 73% year-over-year. The full fiscal year revenue was $215.9 billion, a 65% increase compared to the previous year.

In Q4, NVIDIA’s GAAP and non-GAAP gross margins were 75.0% and 75.2%, respectively. For the full year, GAAP and non-GAAP gross margins were 71.1% and 71.3%.

“The AI agent inflection point has arrived”

“Computing demand is growing exponentially, and the inflection point for AI agents has arrived. Thanks to NVLink technology, NVIDIA’s Grace Blackwell chips have reduced inference costs per token by an order of magnitude, and the next-generation Vera Rubin chips will further solidify the company’s leadership,” said Jensen Huang, founder and CEO of NVIDIA. “Enterprise adoption of AI agents is rapidly increasing, with customers rushing to invest in AI computing power. NVIDIA’s computing factories will further drive the industrial revolution brought by AI and support future growth for businesses.”

Previously, NVIDIA provided a Q4 revenue guidance of $65 billion, with a variance of ±2%; GAAP and non-GAAP gross margins were expected to be 74.8% and 75.0%, respectively, with a 50 basis point variance. The market widely expected the company’s Q4 revenue to reach $66.2 billion. Overall, NVIDIA’s financial performance in Q4 of FY2026 has exceeded previous expectations.

Industry analysts believe that NVIDIA’s strong performance is not particularly surprising, and this growth momentum may continue for some time. Recently, cloud service providers like Alphabet and Amazon reported quarterly results. It is estimated that the combined capital expenditure of the four major cloud providers (Alphabet, Amazon, Meta, and Microsoft) could approach $700 billion this year, with a significant portion likely flowing directly to NVIDIA. Additionally, NVIDIA has sold high-priced GPU chips to major tech companies and high-valued AI startups.

Revenue may continue to grow in the next quarter

Specifically, NVIDIA’s business segments:

Data Center: Benefiting from growth in accelerated computing and AI, Q4 revenue reached a record $62.3 billion, up 22% quarter-over-quarter and 75% year-over-year; full-year revenue increased 68% year-over-year to a record $193.7 billion.

Gaming and AI PCs: Driven by strong demand for Blackwell products, Q4 revenue was $3.7 billion, up 47% year-over-year; however, channel inventory naturally declined after the holiday season, resulting in a 13% decrease quarter-over-quarter. Full-year revenue grew 41% year-over-year to a record $16 billion.

Professional Visualization: Q4 revenue was $1.3 billion, up 74% quarter-over-quarter and 159% year-over-year; full-year revenue increased 70% year-over-year to a record $3.2 billion.

Automotive and Robotics: Thanks to the continued adoption of autonomous driving, Q4 revenue was $604 million, up 2% quarter-over-quarter and 6% year-over-year; full-year revenue reached a record $2.3 billion, up 39% year-over-year.

NVIDIA’s outlook for Q1 FY2027 is revenue of $78 billion, with a variance of ±2%; GAAP and non-GAAP gross margins are expected to be 74.9% and 75.0%, respectively, with a 50 basis point variance. Notably, the outlook does not include revenue from China’s data center computing business.

Vera Rubin planned for mass production in the second half of the year

NVIDIA’s earnings call indicated that supply of cutting-edge chips remains tight, and embodied and physical AI applications are beginning to drive revenue growth.

According to NVIDIA, the Grace Blackwell system accounts for two-thirds of data center revenue in Q4, with outstanding network business performance and record-high demand for large-scale deployment and technological expansion. The company expects growth in sovereign computing to stay in sync with its AI infrastructure business, with confidence in cash flow growth from major clients. NVIDIA plans to leverage new technologies from Groq to expand its product line.

“Currently, we are still negotiating investment terms with OpenAI, and no agreement has been finalized,” Huang said during the earnings presentation.

Market expectations for NVIDIA’s next-generation rack system, Vera Rubin, are high. CFO Colette Kress stated during the earnings call: “Earlier this week, we shipped the first batch of Vera Rubin prototypes to customers, and mass production is still on track to begin in the second half of this year.”

Vera Rubin is expected to deliver a 10x performance increase per watt, providing higher energy efficiency amid severe power constraints in data centers. The company is expanding supply chains from Asia to the U.S. and Latin America. NVIDIA says these measures are expected to strengthen supply chain resilience and redundancy to meet the growing demand for AI infrastructure. The success of increasing manufacturing capacity will depend on the local manufacturing ecosystem’s ability to scale up production in time.

Dion Harris, head of NVIDIA’s AI infrastructure, showcased the internal configuration and supplier details of the complete Vera Rubin rack in a recent media interview. Besides the 72 Rubin GPUs and 36 CPUs in the core, the entire rack contains 1.3 million components supplied by over 80 vendors from more than 20 countries, including China, Vietnam, and Thailand.

Due to increased power consumption, Vera Rubin is NVIDIA’s first system to be 100% liquid-cooled. NVIDIA has advised customers that future AI factories will predominantly adopt liquid cooling architectures. The closed-loop liquid cooling design also conserves water resources.