Three years ago, Nvidia’s annual profit was $4.4 billion. Last year, it crossed $120 billion.
The company said Wednesday that revenue for its fiscal fourth quarter, which ended January 25, climbed to $68.1 billion, up 73% from a year earlier. Net income for the quarter nearly doubled to $43 billion, or $1.76 per share. For the full fiscal year, profit reached $120 billion.
The real engine behind Nvidia’s performance remains its data center division, which houses its powerful AI chips. Fourth-quarter data center revenue hit a record $62.3 billion, up 75% year over year. For the full fiscal year, data center revenue rose 68% to $193.7 billion, according to Nvidia’s earnings release.
In a statement accompanying the results, CEO Jensen Huang pointed to explosive demand for computing power.
“Computing demand is growing exponentially — the agentic AI inflection point has arrived. Grace Blackwell with NVLink is the king of inference today — delivering an order-of-magnitude lower cost per token — and Vera Rubin will extend that leadership even further,” Huang said in Nvidia’s earnings release. “Our customers are racing to invest in AI compute — the factories powering the AI industrial revolution and their future growth.”
Nvidia also issued stronger-than-expected guidance for the current quarter. The company forecast revenue of $78 billion, plus or minus 2%, well above analyst estimates of $72 to $73 billion. Notably, Nvidia said it is not assuming any data center revenue from China in its outlook. Shares rose in after-hours trading immediately after the report, though gains later narrowed.
Hyperscalers lead spending surge
Much of Nvidia’s growth continues to come from major cloud providers.
Hyperscalers accounted for just over 50% of data center revenue, according to CFO Colette Kress. “For the fourth quarter, hyperscaler revenue increased and remained our largest customer category at slightly over 50% of Data Center revenue, while growth was led by the rest of our Data Center customers as revenue diversified,” Kress said in a statement.
Within data centers, networking revenue jumped 263% year over year to nearly $11 billion, reflecting growing adoption of Nvidia’s NVLink and Spectrum-X technologies.
Gaming slips, but don’t panic
Nvidia’s gaming division, which used to be its bread and butter, brought in $3.7 billion for the quarter. That’s up 47% from last year but down 13% from the previous quarter. The company blamed “naturally moderated” channel inventory following the holiday season.
There might be some turbulence ahead for gaming, though. Kress noted in her commentary that the company expects supply constraints to hit the Gaming business “in the first quarter of fiscal 2027 and beyond.” Memory shortages are reportedly forcing chipmakers to prioritize AI processors over gaming GPUs.
The Rubin era is coming
Nvidia isn’t resting on the success of its current Blackwell chips. The company is already moving toward its next big thing: the Vera Rubin platform.
Kress confirmed that the company “shipped our first Vera Rubin samples to customers earlier this week,” with full production expected later this year. These new chips are designed to be 10x more efficient, which is crucial given data centers’ massive power demands.
It’s not all perfectly smooth sailing. While hyperscalers account for about half of Nvidia’s data center revenue, the company is still navigating a tricky geopolitical landscape.
The US government has strictly limited what Nvidia can sell to China. During the earnings call, Kress noted that while they have approval to sell some H200 products there, it’s only in “small amounts,” and it’s unclear if Chinese companies will even be allowed to buy them.
The competition is only intensifying. AMD recently struck massive deals with Meta and OpenAI, and some analysts wonder if Nvidia can keep up with the sheer volume of demand.
