Nvidia’s market capitalisation neared a record $5 trillion (£3.8bn) on Tuesday following several business announcements during its GTC event in Washington, DC, amidst concerns that investors’ enthusiasm around artificial intelligence may not be sustainable.
The AI chipmaker’s shares closed at $4.89tn after rising as high as $4.94tn during the day.
On the same day, Microsoft’s market capitalisation closed above $4tn after it reached a definitive deal with OpenAI over the future of their collaboration, while Apple’s shares briefly rose above the same level due to optimism based on strong iPhone sales, before closing below the mark.
Supercomputers, robotaxis
Nvidia’s announcements at the event included a $1bn investment in Finland’s Nokia that drove its shares up the most since 2013.
Nvidia also said it would build seven supercomputers for the US government, following a similar government supercomputer announcement from AMD.
Electric carmaker Lucid said it would adopt Nvidia’s autonomous driving platform to develop future Level 4 highly automated vehicles, while Stellantis said it would work with Nvidia and others on future autonomous taxis.
Stellantis, which owns brands including Jeep, Ram, Dodge, Chrysler, Fiat and Renault, said it would use Nvidia’s Drive AGX Hyperion 10 platform, which includes its DriveOS software, with hardware and systems integration from Taiwan’s Foxconn.
Production for vehicles with the new technology is scheduled to begin in 2028, and Uber said it would deploy 5,000 of the automated taxis in the US, followed by further vehicles in other markets, without specifying a timeline.
Uber is working with numerous robotaxi companies, including Wayve and China’s Pony.ai and WeRide, while Stellantis is also working with other partners on robotaxis, including Pony.ai.
Unproven technology
Industry and financial analysts have grown increasingly concerned about ambitious financial projections and frothy corporate valuations that they say are based on an unproven technology.
Gene Munster, co-founder of Deepwater Asset Management, noted that a Qualcomm announcement that it is planning AI chips to compete with Nvidia resulted in a substantial stock price rise, even though it has yet to release such a product and faces intense competition.
“What we saw today, there is no substance behind this. This is a press release that has unlocked $20 billion (£15bn) in market cap,” Munster told CNBC’s Squawk Box.
Analysts have noted that in spite of ongoing corporate enthusiasm for AI deployments, companies have only been testing out such technologies for about three years, and there is no proof as yet that they will be widely effective.
Return on investment
A widely cited MIT study in August found that only about 5 percent of AI pilot programmes achieve rapid return on investment, with 95 percent of organisations finding no measurable return on their generative AI investments.
In June, Gartner found that 40 percent of AI agent projects are likely to be cancelled within only two years due to rising costs and a poor return on investment, adding that the use cases for agentic AI is currently immature.
Much agentic AI is driven by hype, with the vast majority of so-called AI agents actually being repackaged versions of existing products such as chatbots, the analysts said.
