AI (artificial intelligence) chip leader Nvidia's revenue exceeded expectations last quarter and gave strong performance guidance. After the U.S. stock market closed on February 26, local time, Nvidia (Nasdaq: NVDA) released its financial report for the fourth quarter of fiscal year 2025 as of January 26. During the period, it achieved revenue of US$39.331 billion, a year-on-year increase of 78%, higher than market expectations of 38 50 million U.S. dollars; net profit under U.S. GAAP was $22.091 billion, a year-on-year increase of 80%; non-GAAP diluted earnings per share was $0.89, a year-on-year increase of 82%, higher than market expectations of $0.84.
Throughout fiscal year 2025, Nvidia’s revenue reached US$130.5 billion, a year-on-year increase of 114%. Non-GAAP diluted earnings per share were $2.99, a year-on-year increase of 130%. The company spent $33.7 billion on stock repurchases in fiscal 2025.
At the same time, NVIDIA has provided guidance for the next quarter's performance, predicting that sales in the first quarter of fiscal year 2026 will reach US$43 billion, plus or minus 2%, which means a year-on-year growth of 65%, higher than the market average expectation of US$41.78 billion. Nvidia will maintain its quarterly dividend of 1 cent per share.
NVIDIA CEO Jensen Huang said: "The demand for Blackwell chips is amazing, because inference AI adds another scaling law - increasing the computing power during training will make the model smarter, and increasing the computing power invested in deep thinking will make the answers smarter. We have successfully achieved mass production of the Blackwell AI supercomputer, achieving billions of dollars in sales in its first quarter of availability. With agent AI and physical AI laying the foundation for the next wave of AI revolution, AI is developing at the speed of light and will change some of the biggest industries."
Summary of Nvidia's fourth quarter fiscal 2025 results. Source: NVIDIA financial report
On the 26th, Nvidia (Nasdaq: NVDA) stock price closed at $131.28 per share, up 3.67%, with a total market value of $3.22 trillion. After the financial report was released after the market closed, NVIDIA's stock price fluctuated violently after the market closed. It once fell by 2%. It turned higher after the earnings call began, but then fell again. As of press time, the company's stock price fell slightly.
It is worth noting that in the fourth fiscal quarter, the company's gross profit margin reached 73.0%, a year-on-year decrease of 3.0% and a month-on-month decrease of 1.6%. Nvidia said the decline in gross profit margins was due to more complex and more expensive new products in the data center business, by far the company's largest source of revenue. After Blackwell enters full production, gross profit margin is expected to increase to around 75% later in fiscal 2026.
From a business perspective, NVIDIA's most watched data center business once again hit a record high. NVIDIA's data center business revenue in the fourth fiscal quarter was US$35.6 billion, higher than market expectations of US$33.65 billion, an increase of 93% compared with the same period last year and a 16% month-on-month increase.
Nvidia CFO Colette Kress said that sales of the latest AI chip Blackwell "exceeded expectations" and brought in $11 billion in revenue in the fourth fiscal quarter. It is the fastest-growing product in the company's history and its production is "at full speed": "Large cloud service providers lead the sales of Blackwell chips, and they account for about 50% of our data center revenue."
Kress emphasized: "The scale of post-training (AI model) and model customization is very large, and driven by test time expansion and new inference models such as OpenAI's o3, DeepSeekR1 and Grok3, it may require greater orders of magnitude computational acceleration."
Huang added that the Blackwell product line is expected to "achieve incredible results in every aspect": "We have to continue to scale because the demand is quite high and customers are eager to get Blackwell systems." The company believes that it expects to achieve strong sequential growth in the first quarter as supply increases and restrictions ease.
However, Kress pointed out that the company's data center revenue in China is "well below expectations" and product shipments will be "roughly maintained at current proportions", while revenue from the U.S. data center business is strongest driven by Blackwell chips. Additionally, the impact of future tariffs remains an unknown.
Huang Renxun also said on the call that the revenue share of China's business in this fiscal quarter is expected to be roughly the same as in the fourth fiscal quarter and before, "about half of what it was before the implementation of export controls." He added that it is best to think of AI chips as software that is "ubiquitous in every country" and that "every student will use AI as a tutor."
Regarding the company's AI chip roadmap, Huang admitted that switching from the previous generation Hopper architecture to the Blackwell architecture was a "very challenging transition." The company encountered "small problems" during the early production of Blackwell and "may have wasted a few months." However, Nvidia recovered at "light speed" and the transition to Blackwell Ultra is expected to be smoother, and it will be available in the second half of this year.
In terms of other businesses, as Nvidia's "old business", the gaming and AIPC (artificial intelligence personal computer) business achieved revenue of US$2.5 billion in the third fiscal quarter, a 22% decrease compared with the same period last year. In addition, professional visualization business revenue was US$511 million, an increase of 10% compared with the same period last year and an increase of 5% compared with the previous fiscal quarter. Automotive and robotics business revenue was US$570 million, an increase of 103% compared with the same period last year and an increase of 27% compared with the previous fiscal quarter.
Previously, on January 20, Chinese AI startup DeepSeek launched the cost-effective and high-performance large-scale open source model DeepSeek-R1, which subsequently caused a sensation in the international market, causing Nvidia's market value to evaporate by US$600 billion in just three days from January 24.
In the final summary of the post-earnings conference call, Huang Renxun said that DeepSeek-R1 "ignited global enthusiasm" and that the inference model will promote the development of the chip business: "
In an interview program broadcast recently, Huang Renxun called DeepSeek's open source inference model R1 "extremely exciting." However, there are some misunderstandings in the market's response to R1. The release of R1 is essentially good for the AI market and can accelerate the adoption of AI, which means that the market still needs computing resources.
Currently, the market is paying close attention to whether large technology companies will continue to invest heavily in AI infrastructure construction and whether the investment can bring corresponding returns. According to reports from Microsoft, Google, Amazon and Meta, the total capital expenditures of the four giants in 2024 will be US$246 billion, a year-on-year surge of 63%, and the total capital expenditures in 2025 may exceed US$320 billion.
According to forecasts released by Morgan Stanley analysts in February this year, Microsoft will account for nearly 35% of the total revenue of Nvidia's latest AI chip Blackwell in 2025, Google will account for 32.2%, Oracle will account for 7.4%, and Amazon will account for 6.2%.
During the conference call, faced with the question of whether Nvidia’s business would be threatened by self-developed customized chips by companies such as Amazon, Microsoft and Google, Huang Renxun said: “Even if the chip is designed, it does not mean that it can be deployed.”
Factset data shows that before the financial report was released, 62 of 68 analysts gave Nvidia a "buy" or "outperform" rating, and the remaining 6 gave it a "hold" rating. Analysts give Nvidia an average price target of $175.39.