Nvidia reported stellar earnings and upbeat forecasts on Wednesday, as Chief Executive Jensen Huang sought to allay concerns about the sustainability of the AI ​​boom that has propelled the chipmaker to the top of the world's most valuable company. The group, which is now comparable to economic data and can influence the sentiment of global stock markets, released its latest results: Nvidia expects revenue for this quarter to reach US$78 billion, exceeding the forecast of US$72.1 billion, and profits surged to a record US$120 billion last year.

Huang Renxun praised Nvidia's performance, saying: "Computing demand is growing exponentially... Our customers are rushing to invest in AI computing power."

However, Nvidia's performance did not wow investors as concerns on Wall Street grew about whether the AI ​​craze might turn into a bubble.

The company's shares initially rose 4% in after-hours trading Wednesday before giving up gains. Shares are expected to open flat on Thursday.

Nvidia is seen as a bellwether in the AI ​​industry because its chips play a key role in running the technology — a status that also makes it one of the most notable winners from the boom.

The results come as Nvidia grapples with growing investor concerns about huge spending by its largest customer in Silicon Valley and a shortage of memory chips that is squeezing the entire industry.

Analysts at TD Cowen noted: "All systems are 'fine' at Nvidia, but skepticism continues to pervade large-cap AI stocks."

After years of breathless gains that pushed Nvidia's market value to $4.7 trillion, Nvidia shares are up just 5% this year.

In its financial report, the company announced revenue of US$68.1 billion in the last quarter of its fiscal year (ending in late January), a year-on-year increase of 73%, exceeding previous guidance of approximately US$65 billion. Results beat analyst expectations compiled by Visible Alpha of $66.2 billion.

Data center revenue - which refers to Nvidia's AI chip technology, including the current generation of Blackwell hardware launched last year - was $62.3 billion, better than expectations of $60.5 billion.

NVIDIA's total revenue for the year exceeded US$200 billion, and net profit in the last quarter was US$43 billion, exceeding expectations of US$36.4 billion.

The company's gross profit margin also met analysts' expectations of 75%, and it said it would maintain that level in the April quarter and maintain profit margins of "more than 70 percent" for the rest of the year.

Despite the strong results, Visible Alpha research director Melissa Otto said Nvidia faces "a lot of concerns" about how the large tech companies and startup customers it relies on will fund hundreds of billions of dollars in AI infrastructure spending.

"These are the key questions for the stock that really explain why it's not rising," she said. "We're not getting a lot of visibility into how the rest of the year is going to play out."

Analysts are increasingly concerned about Nvidia's reliance on a handful of large data center operators known as "hyperscale cloud providers," as well as AI startups such as OpenAI.

Microsoft, Google, Amazon and Meta have a combined expected capital expenditure of US$660 billion this year, mainly focused on AI data centers. The tech giants now face the prospect of tapping debt and stock markets to raise capital while burning through mountains of cash.

AI startups—which are huge consumers of data center capacity for training and running systems like ChatGPT and Anthropic’s Claude—also need to continue to secure huge funding deals to pay for computing power.

Huang dismissed concerns that the pace of spending was unsustainable, saying he was "confident in (hyperscale cloud providers') cash flow growth."

Huang Renxun said that increasing computing power will bring revenue to these groups. “In this new world of AI, computing power equals income.”

He also predicts that AI use will continue to accelerate as advances by cutting-edge AI model builders such as Anthropic and OpenAI "open the floodgates for enterprise adoption of AI."

Previously, it was reported that Nvidia was investing US$30 billion in OpenAI’s latest round of financing, replacing an earlier US$100 billion multi-year strategic partnership that would have been invested in installments as its chips were used.

Huang Renxun said that Nvidia is "excited about working with OpenAI", calling OpenAI a "company where talents appear once in a generation."

Another headwind facing Nvidia is a widespread shortage of memory chips, as data center builders are bulk-buying limited supplies of high-bandwidth memory from a handful of companies — SK Hynix, Samsung and Micron.

Visible Alpha's Otto said the ongoing memory chip supply squeeze means Nvidia's profit margins are now harder to predict because Nvidia needs memory to run with its advanced processors.

Nvidia Chief Financial Officer Colette Kress said the company "has inventory and supply commitments to meet future demand, including shipments extending into calendar year 2027."

Daniel Newman, CEO of The Futurum Group, said that while Nvidia's headline numbers were "perfect," the company faced "thinly veiled suspicion that the show simply couldn't continue."