Home / Business and Politics / Nvidia Breaks the Fear of the AI Bubble: Revenues Soar to $57 Billion

Nvidia Breaks the Fear of the AI Bubble: Revenues Soar to $57 Billion

Nvidia, kompanija, tehnologija
Nvidia, kompanija, tehnologija / Image by: foto Shutterstock

Nvidia has once again exceeded Wall Street expectations in the latest quarter, increasing chip sales faster than the market had predicted, calming investors who feared that the euphoria surrounding artificial intelligence was beginning to deflate.

CEO Jensen Huang dismissed the nervousness surrounding a potential ‘AI bubble’ burst, after hundreds of billions of dollars in market value evaporated from the tech sector in recent weeks.

– There is a lot of talk about the AI bubble. From our perspective, we see something entirely different – Huang told analysts, as reported by the FT.

The world’s most valuable company reported that its revenues for the quarter ending in late October rose 62 percent to $57 billion, above the consensus estimate of analysts of $55 billion. For the current quarter, Nvidia forecasts $65 billion in revenue, which is about $3 billion higher than Wall Street expectations.

Shares rose about 5 percent, and the positive tone of the results immediately spilled over into broader markets, with Asian exchanges opening in the green, Japan’s Nikkei 225 rising 3.7 percent, and South Korea’s Kospi up 2.2 percent.

Nvidia’s results are viewed as a kind of ‘barometer’ for the health of the AI sector, as its advanced chips power the most sophisticated models, including OpenAI’s ChatGPT.

In the weeks leading up to the results announcement, tech stocks had suffered due to increasingly loud doubts about the sustainability of high valuations of major U.S. tech groups and their massive investments in chips and data centers. Nvidia’s stock fell 11 percent from its peak in early November, and additional nervousness was heightened by this year’s sale of a $5.8 billion stock package by SoftBank, as well as warnings from well-known ‘short seller’ Michael Burry about the AI bubble.

The Hard Core of the AI Ecosystem

This year, Nvidia has tied its fate to its largest clients through a series of multi-billion dollar contracts. The company’s market value reached about $4.5 trillion, and some of these contracts include up to $100 billion in planned investments in OpenAI.

Some analysts are increasingly concerned about the ‘closed loop’ of business between a narrow circle of chip manufacturers, AI developers, and cloud service providers, where the same companies simultaneously invest in each other and purchase products from one another.

Despite this, Nvidia’s strong results have lifted the rest of the tech sector, with shares of competitor AMD rising about 4 percent in after-hours trading, and gains recorded by Alphabet, Amazon, Meta, Oracle, and Microsoft.

Daniel Newman, director of the research firm The Futurum Group, assessed that the results confirm that the momentum around AI is ‘still intact’.

– While it is hard to believe that demand is this stable and strong, at some point skeptics will have to start believing – he commented.

Capital and Energy Become Bottlenecks

In a regulatory filing, Nvidia warned for the first time that its clients’ ability to secure capital and energy for AI data centers could become a growth bottleneck. The company states that customers with less financial strength ‘may have difficulty securing financing for large infrastructure projects’, which could lead to delays and slower adoption of AI.

In other words, even if demand for chips is enormous, the bottlenecks are money and electricity.

Nevertheless, Nvidia continues to raise its own forecasts. Its market capitalization briefly exceeded $5 trillion at the end of October, after Huang revealed that the company had already ‘locked in’ over $500 billion in orders for its latest chips until the end of next year.

CFO Colette Kress added that future revenues could grow even more, referring to a $10 billion investment agreement with Anthropic, under which the AI startup will use Nvidia’s chips for the first time.

Saudi Arabia, UAE, and U.S. Export Licenses

At the U.S.-Saudi investment forum on Wednesday, Huang announced that Musk’s company xAI is joining Nvidia’s existing partnership with the Saudi state AI fund Humain. Together, they are working on a new 500-megawatt data center in Saudi Arabia, based on Nvidia’s chips.

On the same day, the U.S. Department of Commerce announced that it had approved the export of Nvidia’s Blackwell chips to Humain and G42, a leading AI company from the United Arab Emirates.

The approval for the purchase of up to 35,000 chips came despite concerns from some U.S. security officials that China could access advanced U.S. AI technology through the UAE.

Growth Despite China

Nvidia’s growth is occurring despite the fact that the company has practically been shut out of the Chinese data center market after a tumultuous year in U.S.-China trade relations.

Nvidia stated that sales of the H20 chip, which was specifically tailored for the Chinese market to comply with U.S. export controls, are ‘insignificant’ due to a combination of tightened U.S. restrictions and pressure from the Chinese government on domestic companies to use local alternative chips. The company currently does not include potential revenues from China in its forecasts.

Net income in the latest quarter rose 65 percent to $31.9 billion, above analysts’ expectations of $30 billion. The gross margin was a very high 73.4 percent.

Revenue from data centers, or from the sale of AI chips, reached $51.2 billion, above forecasts of $49 billion. Nvidia reported ‘continuously strong demand’ for last year’s generation of Blackwell chips, even as it began delivering their successor, Blackwell Ultra, in significant quantities.

The company expects the gross margin in the current quarter to reach the planned 75 percent, after being briefly under pressure during 2025 due to the costs of launching Blackwell technology.

Kress stated that Nvidia is ‘working to maintain gross margin’ at approximately that level in 2026, even as its next generation of chips, codenamed Vera Rubin, comes to market.

Tagged: