Written by Stephen Nellis and Max A. Charney
SAN FRANCISCO (Reuters) – How long can Nvidia and its CEO Jensen Huang hold the crown as the technology world's leading supplier of artificial intelligence chips?
Whether it's months, years or even decades, Huang, who co-founded Nvidia in 1993 and has been its CEO ever since, took the stage at a Silicon Valley hockey arena Monday to unveil a new flagship product. This issue will be key when stepping into.
The occasion is the company's annual developer conference, which will be held in-person for the first time since the pandemic. Nvidia expects 16,000 people to attend the GTC event live. That's nearly double the number of people who attended the 2019 show in person.
Nvidia's market capitalization topped $2 trillion in late February, leaving it “only” $400 billion away from overtaking Apple as Wall Street's second most valuable company after stock market leader Microsoft. There is.
Analysts say NVIDIA's sales will rise 81% to $110 billion this year as the company adds tens of thousands of top-end chips to power chatbots, image generators and other AI systems. I expect it to be.
That's double the revenue analysts expect for Intel, the chipmaker that dominated the personal computer boom. The question for Nvidia is whether it can leverage its recent significant lead in AI computing to long-term advantage in a new era of computing.
At the center of that effort will be Nvidia's next-generation high-end AI processor, which analysts expect to be called the B100. The chip will be at the heart of an AI system sold by Nvidia, and will likely start shipping later this year.
Demand for Nvidia's current AI chips exceeds supply, and software developers have been waiting months for the chance to use AI-optimized computers at cloud providers.
Nvidia is unlikely to reveal specific pricing, but the B100 will likely cost more than its predecessor, which sells for more than $20,000.
Nvidia stock is up 83% in 2024 after more than tripling last year. Even after that meteoric rise, NVIDIA still trades at 34 times forward earnings, according to LSEG data, a discount compared to 58 times forward earnings a year ago.
This drop in Nvidia's P/E valuation is the result of analysts significantly raising their future earnings forecasts for the company, and if those forecasts prove too optimistic, Nvidia's stock price will plummet back down. There is a risk.
Shares fell 1.1% on Wednesday, closing at $908.88.
Bernstein analyst Stacey Rasgon said of Nvidia's recent financial performance, “My biggest concern is that the numbers are growing too quickly. I'm worried that these numbers won't be sustainable.” Just do it.'' “The more new, high-performance, high-priced products we have, the wider the runway.”
Nvidia is also likely to introduce a number of updates to the software known as CUDA, which provides tools for developers to run AI programs on the company's chips, and to lock developers into Nvidia's chips. Helpful. CUDA makes it difficult to switch to components sold by rivals such as Advanced Mirco Devices or proprietary chips from Microsoft and Alphabet's Google.
Nvidia began offering its chips and software as a cloud service to developers last year, and analysts are watching to see how the company expands its efforts.
“The service side is really a software role,” said Ryan Shrout of chip analysis firm Shrout Research. He said analysts will be watching closely to see if cloud and software providers “may be concerned about Nvidia playing in their playground.”
China business
Another key question is whether Santa Clara, Calif.-based Nvidia can establish a decisive technological advantage over its Chinese rivals. The U.S. government has blocked China's access to Nvidia's cutting-edge chips, and China responded with a Huawei chip that rivals the performance of Nvidia's A100 chip, released in 2020.
No Chinese company has yet claimed to be a match for Nvidia's current flagship chip, the H100, due out in 2022, and China watchers expect the upcoming B100 to set China back even further.
“When the government developed these controls, it knew very well that companies like NVIDIA would improve their silicon, software coding, and application stack almost every year,” said Jimmy Goodrich, a China and semiconductor expert. said. “Over time, that difference will grow exponentially.”
(Reporting by Stephen Nellis and Max A. Charney in San Francisco; Editing by Matthew Lewis)