Amidst recent product launches from Anthropic and Citrini's "doomsday report" fueling investor panic, the artificial intelligence (AI) boom has withstood direct scrutiny, with Nvidia delivering phenomenal results demonstrating the continued strong demand generated by AI.

On Wednesday, January 25th, Eastern Time, Nvidia announced that its revenue for the fourth fiscal quarter of 2026 ("Q4"), ending January 31, 2026, reached a record $68.1 billion, an increase of about 70% year-over-year. The core data center business, which contributed more than 90% of the revenue, also set a new quarterly revenue record, both exceeding analysts' expectations by more than 3%.

Nvidia's earnings also performed strongly in the fourth quarter. On a non-GAAP basis, adjusted earnings per share (EPS) increased by more than 80% year-over-year, about 5.9% higher than analysts' expectations, and gross margin also exceeded expectations, climbing to 75.2%, a new high in a year and a half.

Even more encouraging for investors was Nvidia's stronger-than-expected guidance for the first fiscal quarter of 2027. Revenue is expected to reach a new high, with the midpoint of the guidance range 7.1% higher than the median analyst estimate and even 4% higher than the optimistic expectations of buy-side analysts. Year-on-year growth accelerated 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 Wednesday's earnings call, Nvidia CEO Jensen Huang raised its previously stated chip revenue forecast, saying, "We will exceed the $500 billion target." Supply will meet demand for the coming year. At the GTC conference last October, Huang revealed that Nvidia had secured a total of $500 billion in chip orders for 2025 and 2026, including the next-generation Rubin chip, which will begin mass production this year.

Jensen Huang stated that customers are scrambling to invest in AI computing. Computing demand is growing rapidly. Enterprise adoption of agents is skyrocketing. Regarding "space data centers," he said the current economics are "barren," but the situation will change over time.

Following the earnings release, Nvidia's stock price, which had already closed up over 1% on Wednesday, surged in after-hours trading, with gains rapidly expanding to over 4% at one point. Analysts believe the key factors driving the market's positive reaction are: both data center revenue and total revenue exceeded expectations; gross margin continued to improve as production of the new-generation Blackwell architecture chips ramped up; and the guidance for this quarter was stronger, even excluding some revenue from the Chinese market, reinforcing the narrative of resilient AI computing power demand.

However, during the conference call, Nvidia's stock price continued to give back its gains, turning negative in after-hours trading, at one point falling by more than 1%. Some commentators suggested that the stock price decline indicated investors were not swayed by the latest guidance, suggesting that market concerns about an overheated AI economy will continue to plague Nvidia. Other analysts stated that continued high growth in operating expenses and the inclusion of stock-based compensation (SBC) in non-GAAP metrics starting in the first quarter may change investors' perception of "profit growth" in the short term.

Q4 revenue hit a new quarterly high, and gross margin reached a one-and-a-half-year high.

Nvidia's revenue in the fourth quarter increased by 73% year-over-year to $68.127 billion, a significantly higher growth rate than the 62% in the previous quarter, exceeding Nvidia's own guidance midpoint of $65 billion. Analysts had expected revenue of $65.91 billion, representing a year-over-year increase of approximately 68%. Nvidia's full-year revenue also reached a record high of $215.938 billion, a 65% year-over-year increase.

Gross margin was another highlight in 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, a new quarterly high since the second fiscal quarter of fiscal year 2025, higher than the consensus forecast of 74.7% and the optimistic forecast of 75.0%.

Nvidia's Chief Financial Officer (CFO) Colette Kress explained that the year-on-year improvement in gross margin came from "reduced inventory write-offs," while the sequential improvement was related to "a better product and cost structure" resulting from the continued ramp-up of Blackwell chips.

However, throughout fiscal year 2026, the non-GAAP gross margin declined from 75.5% in the previous fiscal year to 71.3%, a year-on-year decrease of 4.2 percentage points, indicating that the annual profit margin will still be subject to structural disturbances during the platform switching and supply ramp-up phase.

Data Centers: Computing Power Growth Stabilizes, Network Takeover Accelerates

Nvidia's data center business reported revenue of $62.314 billion in the fourth quarter, a 75% year-over-year increase, higher than the 66% growth in the previous quarter. Analysts had expected a year-over-year increase of nearly 70% to $60.36 billion.

Inside the data center, Nvidia provided two more noteworthy sets of figures:

• Data center computing (Compute) revenue was $51.334 billion, up 58% year-over-year, slightly higher than the 56% growth rate in the third quarter.
• Data center networking revenue reached $10.98 billion, a year-over-year increase of 263%, far exceeding the 162% growth rate in the third quarter.

NVIDIA attributes the surge in networking revenue to the “launch and continued ramp-up” of the NVLink compute fabric for GB200 and GB300 systems, while Ethernet and InfiniBand platforms continue to grow.

In other words, the market should not only focus on the shipment pace of GPUs themselves, but also see that Nvidia is packaging "computing power, interconnect, and system" into a more irreplaceable overall solution, and the high growth rate of network 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 slightly over 50% of the total revenue of the data center business, remaining the largest customer category. However, the revenue growth in the quarter came more from other data center customers, indicating that revenue sources are diversifying and the risk of concentration is easing.

Blackwell's boost to game demand is subject to short-term supply and distribution disruptions.

Nvidia's gaming revenue reached $3.727 billion in the fourth quarter, a 47% year-over-year increase, compared to analysts' expectations of $4.01 billion and a 30% year-over-year increase in the previous quarter.

Nvidia's gaming business saw accelerated year-over-year growth in the fourth quarter, primarily driven by strong demand for Blackwell chips. However, revenue in this segment declined 13% quarter-over-quarter due to "natural inventory declines following the holiday season." Notably, Nvidia explicitly warned that supply constraints are expected to be a headwind for the gaming business in the first quarter and beyond.

Professional visualization revenue reached $1.321 billion in the fourth quarter, up 159% year-over-year, compared to analysts' expectations of $770.7 million and a 56% year-over-year increase in the previous quarter.

Professional visualization, driven by Blackwell, also saw its revenue more than double year-over-year and grow 74% quarter-over-quarter, becoming one of the most impressive growth businesses outside of data centers. However, this business is still far smaller than that of data centers.

The median revenue guidance for Q1 is expected to increase by nearly 77% year-over-year, excluding revenue from data center computing in China.

Regarding earnings guidance, Nvidia announced that it expects first-quarter revenue of $78 billion, plus or minus 2%, or $76.44 billion to $79.56 billion. This range means that Nvidia's revenue this quarter will surpass the record high set in the fourth quarter.

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

Nvidia's revenue guidance not only exceeded the median analyst estimate of $72.78 billion, but also surpassed the optimistic buy-side estimate of $74 billion to $75 billion.

Nvidia's first-quarter gross margin was in line with optimistic expectations from Wall Street buyers and is expected to reach a new high since the second fiscal quarter of 2025.

The adjusted gross margin for the first quarter on a non-GAAP basis is expected to be 75%, fluctuating by 50 basis points, or 74.5% to 75.5%. The optimistic expectation from buyers is 75%, while the consensus expectation from sellers is 74.7%.

Starting from Q1, non-GAAP incentives were included in equity incentive plans.

Along with its earnings report, Nvidia announced that, starting from the first quarter, non-GAAP financial metrics will no longer exclude stock-based compensation (SBC). Due to this adjustment, Nvidia expects non-GAAP operating expenses to be impacted by approximately $1.9 billion in the first quarter.

This change will directly alter the "usual approach" that the market has long used to compare profit margins and expense ratios. In the short term, it may lead to a recalibration of consensus expectation models and allow investors to see more clearly the real costs that Nvidia has paid to maintain its talent and R&D leadership.