By Chavi Mehta, Stephen Nellis and Jane Lanhee Lee
(Reuters) -Chip designer Nvidia Corp forecast first-quarter revenue above Wall Street estimates on Wednesday as the company benefits from a boom in spending to power chatbots and other artificial intelligence technologies.
The sales outlook sent Nvidia’s shares up 8% in extended trading. Nvidia, the world’s largest supplier of chips used in data centers for training AI, has become a key hardware supplier for large tech companies such as Microsoft Corp that are building services like chat-powered search engines.
Analysts believe Nvidia is best placed to capitalize on this boom as it dominates roughly 80% of the market for graphics processing units, or GPUs.
On a conference call with investors, Nvidia Chief Executive Jensen Huang announced a new service in which Nvidia will offer its own cloud computing service for companies to rent all of its technologies to develop their own “generative” AI services that can create text, images and other forms of data.
AI is still “not deployed in enterprises broadly, but we believe that by hosting everything in the cloud, from the infrastructure through the operating system software, all the way through pre-trained models, we can accelerate the adoption of generative AI in enterprises,” Huang said.
The company forecast current-quarter revenue of $6.50 billion, plus or minus 2%. Analysts on average expect $6.33 billion in revenue, according to Refinitiv data.
Revenue in the quarter ended Jan. 29 was $6.05 billion, compared with analysts’ average estimate of $6.01 billion.
“The launch of generative AI models and the AI arms race taking place should drive accelerated adoption of the company’s new H100 products,” said Logan Purk, an analyst with Edward Jones.
Nvidia’s outlook also helped boost the share prices of competitors such as Advanced Micro Devices, whose shares were up 3% after Nvidia’s results.
The Santa Clara, California, company got its start in the graphics chip business for PCs by helping video games look more realistic. While its revenue beat Wall Street expectations, Nvidia’s sales were still down overall year-on-year as the company weathers a downturn in the PC market.
But growth has remained brisk in the market for data center chips, which has become as important to Nvidia as gaming. The company dominates the business of training artificial intelligence models, with its chips making up about 86% of the computing power of the world’s six biggest cloud services as of December, according to Jefferies chip analyst Mark Lipacis.
Research firm Gartner expects revenue from GPUs used for AI work alone to grow to $12.17 billion by 2026 from $5.9 billion last year.
Last quarter, Nvidia said tech giant Microsoft was deploying “tens of thousands” of its GPUs in its cloud unit Azure to train and use AI. Microsoft-backed OpenAI’s ChatGPT runs entirely on Nvidia chips, Piper Sandler analyst Harsh Kumar said. Analysts at Bank of America Global research believe the boom in so-called generative AI like chatbot and image creation services could add $14 billion more to Nvidia’s revenue by 2027.
On a conference call, Huang said that the use of Nvidia’s chips to run what are called large language models – the technology behind services like ChatGPT – “has just gone through the roof in the last 60 days.”
Revenue from the data center business was $3.62 billion for the fourth quarter, slightly below analyst estimates of $3.84 billion. Gaming chip sales were $1.83 billion, beating analyst estimates of $1.52 billion, according to Refinitiv data.
Adjusted profit was 88 cents per share for the fourth quarter, beating analyst estimates of 81 cents.
(Reporting by Chavi Mehta in Bengaluru, Stephen Nellis in San Francisco and Jane Lee in Oakland, Calif.Editing by Shinjini Ganguli, Peter Henderson and Matthew Lewis)
