Nvidia beats revenue expectations amid DeepSeek fears
Make us a preferred source
Make us a preferred source on Google
The news: Nvidia posted a 12% rise in revenue to USD39.33 billion ($62.43 billion) for the quarter ending January 26, coming ahead of analyst expectations despite advances from China's DeepSeek lab seeding doubts about AI infrastructure growth.
CEO Jensen Huang said the numbers were buoyed by demand for its new Blackwell chips, which are fuelling the development of next-generation "reasoning" AI models.
The numbers: Earnings per share came in at USD0.89 per share, ahead of analyst expectations of USD0.84. Data centre revenue, Nvidia’s biggest segment, reached USD35.6 billion, a 93% increase from last year and a 16% jump for the quarter ending November.
Nvidia expects revenue of USD43 billion ($68 billion) for the current quarter. Net income for its final quarter of 2025 was USD22 billion ($34 billion), up 14% from the third quarter. Total revenue for the year was up 114% to USD130 billion ($206 billion).
Nvidia's gaming sector, its bread and butter until a few years ago, was down 22% quarter-on-quarter to USD2.5 billion ($3.9 billion).
The context: Nvidia's historic run hit a roadblock in January after Chinese AI lab DeepSeek launched a model competitive with OpenAI's that it says it trained with older, second-grade Nvidia chips for just USD5.5 million.
The idea that AI models can be made dramatically more efficiently spooked Nvidia investors, who reasoned such a development would mean fewer of the tech giant's chips would be needed by the Microsofts and Googles of the world. Nvidia lost nearly USD600 billion of its market cap in a single day, the biggest dive in history, though it's since rebounded.
Huang said that demand for Nvidia's most powerful chips is in fact rising thanks to the advent of "reasoning" AI models, like OpenAI's o1 and o3-mini. These models take longer to answer a query, and expend more computation power, to give a more thoughtful answer.
What they said: Huang said that “Demand for Blackwell is amazing as reasoning AI adds another scaling law — increasing compute for training makes models smarter and increasing compute for long thinking makes the answer smarter."
“We’ve successfully ramped up the massive-scale production of Blackwell AI supercomputers, achieving billions of dollars in sales in its first quarter. AI is advancing at light speed as agentic AI and physical AI set the stage for the next wave of AI to revolutionize the largest industries.”
The source: Nvidia filing
