Amid a series of product releases from Anthropic and Citrini's "end report" exacerbating investor panic, the artificial intelligence (AI) boom withstood direct scrutiny, as Nvidia delivered "shocking" results, proving that demand generated by AI remains robust.

On Wednesday, Eastern Time, February 25, Nvidia announced that for the fourth quarter of its fiscal year 2026 (ending January 31, 2026), revenue reached a record $68.1 billion, a year-on-year increase of approximately 70%, with the core data center business contributing over 90% of revenue and also achieving a new single-quarter income high, exceeding analysts' expectations by over 3%.

In the fourth quarter, Nvidia's profits were equally strong. On a non-GAAP basis, the adjusted earnings per share (EPS) grew by over 80% year-on-year, approximately 5.9% higher than analysts' expectations, and the gross margin exceeded expectations, rising to 75.2%, a new high in a year and a half.

More encouraging for investors was that Nvidia's guidance for the first quarter of fiscal year 2027 also surpassed expectations. Revenue is expected to set new records, with the median of the guidance range 7.1% higher than the analysts' expected median and even 4% higher than bullish expectations, with year-on-year growth accelerating to nearly 77% compared to the fourth quarter. Nvidia noted that this guidance does not include data center computing revenue from the Chinese market.

During the earnings call on Wednesday, Nvidia CEO Jensen Huang also raised his previously stated chip revenue expectations, saying we will exceed the $500 billion target. Supply will meet the demand for the foreseeable future until next year. At the GTC conference last October, Huang revealed that Nvidia had accumulated $500 billion in chip orders for the calendar years 2025 and 2026, including the next-generation Rubin chip that will begin mass production this year.

Huang stated that customers are competing to invest in AI computing. The demand for computing is growing rapidly. Companies are ramping up their applications for agents. He mentioned "space data centers," stating that the current space data center economy is still "barren," but the situation will change over time.

After the earnings report was released, Nvidia's stock price, which had already increased by over 1% on Wednesday, surged further in after-hours trading, at one point rising over 4%. Analysts believe that the key to the market's acceptance lies in: both data center revenue and total revenue exceeding expectations; the gross margin continuing to improve as production of the next-generation Blackwell chips ramps up, and without considering certain revenue from the Chinese market, the guidance for this fiscal quarter is stronger, reinforcing the narrative of resilience in AI computing demand.

However, during the conference call, Nvidia's stock price continuously gave back its gains, turning negative in after-hours trading, with losses exceeding 1%. Some commentators noted that the price decline shows investors were not moved by the latest guidance, indicating that concerns about an overheated AI economy will continue to trouble Nvidia. Other analyses indicated that high growth in operating expenses, along with incorporating stock-based compensation (SBC) into the non-GAAP metrics starting in the first quarter, may alter investors' perceptions of "profit growth" in the short term.

Q4 Revenue Hits Record High, Gross Margin at Year-and-a-Half High

In the fourth quarter, Nvidia's revenue grew by 73% year-on-year to $68.127 billion, with a significant increase compared to the previous quarter's 62%, surpassing Nvidia's own guidance median of $65 billion. Analysts had projected revenue of $65.91 billion, representing a year-on-year increase of about 68%. For the entire fiscal year, Nvidia's revenue also set an annual record at $215.938 billion, up 65% year-on-year.

The gross margin became another highlight of the fourth quarter: the non-GAAP gross margin was 75.2%, up 1.7 percentage points year-on-year and 1.6 percentage points quarter-on-quarter, marking the highest single-quarter level since the second quarter of fiscal year 2025, exceeding the analysts' consensus expectation of 74.7% and the optimistic expectation of 75.0%.

Nvidia's Chief Financial Officer (CFO) Colette Kress explained that the year-on-year improvement in gross margin stemmed from a "reduction in inventory provisions," while the quarter-on-quarter improvement was associated with "better product and cost structures" brought about by the sustained ramp-up of Blackwell chips.

However, for the entire fiscal year 2026, the non-GAAP gross margin declined from the previous fiscal year's 75.5% to 71.3%, reflecting a 4.2 percentage point year-on-year drop, indicating that during the platform transition and supply ramp-up phase, annual profit margins will still be affected by structural disturbances.

Data Centers: Computing Power Growth Stabilizes, Networking Accelerates

In the fourth quarter, Nvidia's data center business recorded revenue of $62.314 billion, a year-on-year increase of 75%, with growth exceeding the previous quarter's 66%, while analysts projected a nearly 70% year-on-year increase to $60.36 billion.

Within the data center segment, Nvidia provided two noteworthy sets of figures:

• Data center computing revenue was $51.334 billion, up 58% year-on-year, a slight acceleration over the third quarter's 56%.
• Data center networking revenue was $10.980 billion, up 263% year-on-year, significantly exceeding the third quarter's 162%.

Nvidia attributed the explosion in networking revenue to the "launch and continuous ramp-up" of NVLink compute fabric for GB200 and GB300 systems, along with the continued growth of Ethernet and InfiniBand platforms.

In other words, the market should not only focus on the shipment pace of GPUs themselves, but should also see that Nvidia is packaging "computing power, interconnects, and systems" into a more irreplaceable overall solution, and the rapid growth in networking revenue is a financial reflection of this strategy.

Regarding the customer structure, the company disclosed that in the fourth quarter, revenue from hyperscale cloud vendors accounted for just over 50% of total revenue in the data center business, remaining the largest customer category, but the revenue growth for the quarter came more from other data center customers, showing a diffusion of revenue sources and a marginal alleviation of concentration risks.

Blackwell Drives Gaming Demand, Short-Term Supply and Channel Disturbances

In the fourth quarter, Nvidia's gaming business revenue was $3.727 billion, a year-on-year increase of 47%, with analysts' expectations set at $4.01 billion, and a year-on-year growth of 30% in the previous quarter.

The gaming business accelerated its growth in the fourth quarter, which Nvidia explained was primarily driven by strong demand for Blackwell chips. However, the revenue for this business decreased by 13% quarter-on-quarter due to "natural declines in channel inventories post-holiday peak." It is worth noting that Nvidia explicitly warned that supply constraints are expected to become a headwind for the gaming business starting in the first quarter and beyond.

Fourth-quarter professional visualization revenue was $1.321 billion, a year-on-year increase of 159%, with analysts expecting $770.7 million, and a year-on-year growth of 56% in the previous quarter.

Professional visualization also achieved more than double revenue growth year-on-year and 74% quarter-on-quarter growth with the support of Blackwell, becoming one of the most impressive incremental businesses besides the data center. However, the size of this business is far less than that of the data center.

Q1 Revenue Guidance Median Year-on-Year Growth of Nearly 77%, Excluding Chinese Data Center Computing Revenue

Regarding guidance, Nvidia announced that first-quarter revenue is expected to be $78 billion, with a fluctuation of 2% around this figure, meaning $76.44 billion to $79.56 billion. This range indicates that Nvidia's revenue in this fiscal quarter will set a new record, surpassing the highest record set in the fourth quarter.

Based on the median of the revenue guidance, Nvidia expects first-quarter revenue to grow year-on-year by 76.9%, further accelerating from the fourth-quarter growth of 73%.

Nvidia's revenue guidance median not only exceeds the analysts’ expected median of $72.78 billion but also surpasses the optimistic expectations of buyers of $74 billion to $75 billion.

Nvidia's first-quarter gross margin aligns with the optimistic expectations of Wall Street buyers, likely setting a new high since the second quarter of fiscal year 2025.

For the first quarter, the non-GAAP adjusted gross margin is expected to be 75%, with a fluctuation of 50 basis points, meaning 74.5% to 75.5%, which matches the buyers' optimistic expectation of 75%, while the sellers' consensus expectation is 74.7%.

Starting Q1, Non-GAAP Will Include Stock-Based Compensation

Along with the financial results announcement, Nvidia stated that starting from the first quarter, financial metrics on a non-GAAP basis will no longer exclude stock-based compensation (SBC). Due to this adjustment, Nvidia expects the impact on operating expenses in the first quarter to be around $1.9 billion.

This change will directly alter the long-standing "conventional metrics" used by the market to compare profit margins and expense ratios, and in the short term, may lead to recalibration of consensus expectation models, allowing investors to more clearly see the true costs Nvidia incurs to retain talent and maintain its research and development leadership.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。