Nvidia clocked a 94 per cent jump in profit to $43bn and pushed fourth-quarter sales up 73 per cent to $68.1bn, putting the cocaine nose job of Wall Street’s panic merchants back in their box.
FactSet’s consensus had net income at $37.5bn on revenue of $66.1bn, so the bar was cleared with room to spare.
Data centre kit made up 91.4 per cent of sales, or $62.3bn, as the company kept feeding the AI and cloud crowd with chips and networking gear.
Nvidia, chief executive Jensen Huang said: “The simple way to think about it is, computing has changed. In this new world of AI, compute equals revenues. I am certain at this point that we’ve reached the point where agentic AI is reshaping business and profits.”
With Nvidia sitting near a $5tn market value, the cocaine nose jobs of Wall Street have got fussy about anything that looks less than flawless. Futurum Group, chief executive Daniel Newman said: “It’s no longer enough for Nvidia to produce good quarterly results. They have to produce perfect quarterly results.”
Gross margins hit 75 per cent in the January quarter, up from 73 per cent a year earlier, landing right where analysts expected. The shares slid to $170.94 in mid-December, then bounced above $196.
Some of Nvidia’s biggest customers include OpenAI, Oracle, Microsoft, Meta Platforms, Alphabet and Amazon, but nerves have built around OpenAI’s fundraising and rival chip efforts, including custom designs.
The Wall Street Journal has reported Nvidia’s planned investment of up to $100bn in OpenAI was put on ice, with people briefed saying a smaller cheque of around $30bn is now in play.
Newman reckons the customer mix has its own risk profile, saying: “They’re exposed to a subset of companies with fragile balance sheets. The next swing factor is the industry’s shift from training AI models to inference, which changes what sort of computing customers buy.”
Nvidia has long owned training with GPUs, but inference leans harder on CPUs, which more firms can design and build around. A recent partnership with Meta highlighted Nvidia’s first major deployment of CPUs not tied to servers, a nod to rising inference demand.
Huang leaned into the money bit, saying: “It’s important to understand that inference equals revenues for our customers now,” and added that customers see Nvidia spend turning into faster growth, Huang said.
Nvidia, chief financial officer Colette Kress brushed off custom-chip threats after the results, saying: “It doesn’t surprise us that others want to take a part of the great market that we’re a big part of. It sounds like a good idea, but right now we’re the king of inference. That’s what we are: the king of inference.”
China is still a messy footnote, even with the Trump administration easing restrictions on sales of H200 chips there. On the call, Kress said: “We have yet to generate any revenue, and we do not know whether any imports will be allowed into China.”
For the current quarter, Nvidia guided to $78bn in revenue, well ahead of the $72.9bn analysts expected, and kept gross margin guidance at 75 per cent.
