After the market close on November 19, 2025 (Eastern Time), NVIDIA, the leading global enterprise in the field of artificial intelligence chips, released its financial report for the third quarter of the 2026 fiscal year. During this period, several key financial indicators significantly exceeded market expectations, and the outlook for the fourth quarter further boosted the confidence of the capital market. After the release of the financial report, NVIDIA's stock price soared by more than 6% in the after-hours trading, and this also led to a simultaneous rise in the stock prices of upstream and downstream enterprises in the semiconductor industry chain.
Both revenue and profit have increased, and the company's profitability remains consistently at the forefront.
In the third fiscal quarter, NVIDIA achieved a total revenue of 57.01 billion US dollars, with a year-on-year growth rate of 62%. This not only significantly exceeded the market expectation of 55.19 billion US dollars, but also surpassed the upper limit of the company's previous performance guidance range. The profit performance was equally outstanding, with a net profit of 31.9 billion US dollars, a year-on-year growth of 65%, and a net profit margin rising to 55.9%. After adjustment, the earnings per share was 1.30 US dollars, 3.1% higher than the analysts' expectation of 1.26 US dollars, further consolidating market confidence.
In terms of gross profit margin, the adjusted gross profit margin for this quarter was 73.6%, which slightly decreased by 1.4 percentage points compared to the previous year and was slightly lower than the market expectation of 73.7%. However, it has rebounded compared to the 72.7% in the previous quarter, highlighting the supporting role of product structure optimization on the gross profit level. The R&D investment has continued to increase, with the R&D expenditure reaching 4.71 billion US dollars this quarter, a year-on-year increase of 39%, providing a solid guarantee for the company's technological iteration and product innovation.
In terms of shareholder returns, the company continues to implement an active dividend policy. In the third fiscal quarter, NVIDIA completed a $12.5 billion stock repurchase and paid out $243 million in dividends. In the first nine months of this fiscal year, the cumulative amount of stock repurchases and dividends reached $37 billion, with remaining stock repurchase capacity at $62.2 billion. By the end of the quarter, the company's cash on hand increased to $60.6 billion, up by $17.4 billion compared to the previous period, and its financial situation remained stable.
The data center business led the growth, while the diversified businesses demonstrated resilience.
The data center business remains NVIDIA's core growth engine. In the third fiscal quarter, this business achieved a record revenue of 51.2 billion US dollars, an increase of 66% year-on-year, accounting for 89% of the total revenue, far exceeding the market expectation of 49.34 billion US dollars. Among them, the computing business (GPU) contributed 43 billion US dollars in revenue, and the network business contributed 8.2 billion US dollars, with the synergy of the two sectors driving business growth.
This outstanding performance is mainly attributed to the strong sales of the Blackwell architecture chips. Nvidia's CEO, Huang Renxun, disclosed that "the sales of the Blackwell series chips exceeded expectations. All cloud GPU products have been sold out", among which the second-generation product, Blackwell Ultra, has become the most popular chip series at present.
The traditional business segments also demonstrated strong resilience in their development. The gaming and AI PC business achieved revenue of 4.3 billion US dollars, representing a year-on-year growth of 30%. Although this was slightly lower than the market expectation of 4.42 billion US dollars, with the update and replacement of the RTX 50 series products and the rapid development of the AI PC market, the foundation for business growth has been continuously strengthened.
The professional visualization business performance exceeded expectations, achieving revenue of 760 million US dollars, a year-on-year increase of 56%, far exceeding the market expectation of 612.8 million US dollars. New products such as DGX Spark AI desktops successfully expanded new application scenarios.
The revenue from the automotive and robotics business reached 592 million US dollars, up by 32% year-on-year. Although it was slightly lower than the market expectation of 621.6 million US dollars, the Orin X chip has successfully entered the supply chains of three new automotive companies, and the intelligent driving chip has begun to enter a stage of scale growth, laying the foundation for the company's long-term development.
The fourth-quarter guidance exceeded expectations, with large orders ensuring long-term growth.
The fourth-quarter earnings outlook announced in the financial report further ignited market enthusiasm. The company expects its total revenue to reach $65 billion, with a fluctuation range of 2% (i.e. $63.7 - $66.3 billion), far exceeding the market expectation of $62 billion. It is projected to have a year-on-year growth rate of over 65%. Profitability is expected to further improve, with the adjusted gross margin expected to reach 75.0% (±50 basis points), an increase of 1.4 percentage points compared to this quarter. The premium effect of Blackwell architecture chips will become more pronounced.
In terms of long-term growth, the company's management reaffirmed that the total order amount for the period 2025-2026 will reach 500 billion US dollars, with the Blackwell and Rubin projects serving as significant supports.
Colette Kress, the Chief Financial Officer of NVIDIA, disclosed that the Rubin chip is expected to enter the ramp-up stage for mass production in the second half of 2026. Meanwhile, the company is collaborating with OpenAI to build an artificial intelligence data center with a total power capacity exceeding 10 gigawatts, and the ecological cooperation is continuously deepening. Kress emphasized that the transformation of ultra-large-scale data centers towards accelerated computing and generative artificial intelligence will contribute half of the long-term growth opportunities, while the physical artificial intelligence market holds trillions of dollars of development potential.
Policy impacts can be controlled, while technical barriers help build competitive advantages.
Currently, Nvidia is still facing uncertainties brought about by the external policy environment. The export control measures imposed by the United States on China have had an impact of approximately 2.5 billion US dollars on the company's revenue. The company has responded by launching the H20 special edition product, but this product has recently encountered sales difficulties in the Chinese market.
It is reported that due to the narrowing gap in performance between the H20 product and domestic chips, as well as safety considerations, there are rumors that the price of the H20 product has dropped significantly and even that it may be discontinued. Currently, the company plans to introduce the B30A chip based on the Blackwell architecture as a replacement to maintain its market share in China.
Regarding the widely concerned issue of "AI bubble", Huang Renxun stated at the analyst meeting, "No signs of an AI bubble have been observed", and emphasized that the computing demands in the training and reasoning fields are growing exponentially. The AI ecosystem is continuously expanding into more industries and countries, providing strong impetus for the company's continuous development.
