Wall Street's expectations were too conservative? NVIDIA Corporation (NVDA.US) CFO stated that revenue will "definitely" exceed previous guidance: over $50 billion!

In response to concerns about the AI spending bubble, NVIDIA Corporation stated that its previously released revenue guidance will become more optimistic due to strong demand. NVIDIA's Chief Financial Officer Colette Kress said at an event hosted by JPMorgan that due to strong demand, NVIDIA is now more bullish on its data center business. By the end of 2026, the expected revenue from NVIDIA's data center chips will "definitely" exceed the forecast of $500 billion given last October. Although the company did not provide a new figure, it hinted that revenue during this period will exceed $500 billion. NVIDIA CEO Jensen Huang said at a press conference, "Our performance this year should be very good. Deals with AI provider Anthropic PBC and improvements in the Chinese market outlook should raise our expectations for this figure." The $500 billion mentioned by Kress refers to the figure Huang mentioned at the GTC conference over two months ago, that by the end of 2026, NVIDIA's existing and future data center chips will generate around $500 billion in revenue. Wall Street currently predicts that NVIDIA's total revenue in 2026 will reach $321.2 billion, a 57% year-on-year increase. They forecast that the company's sales in 2027 will exceed $400 billion. At CES on Tuesday, NVIDIA stated that large customer deals and the increasing popularity of new AI models have resulted in orders exceeding expectations. NVIDIA's outlook is crucial in alleviating investors' concerns about the AI industry. The company, which provides chips for AI models, has always maintained that customer spending will remain strong due to the huge profits generated by AI. AI has become a major engine of global growth. From healthcare diagnostics to financial market analysis, various industries are accelerating their adoption of AI solutions. Cloud service providers such as Amazon, Google, Microsoft, and Meta are significantly increasing capital expenditures specifically for AI infrastructure development. It is estimated that by 2026, the combined capital expenditures of the eight global cloud service providers will reach $600 billion. With over 80% market share in the AI accelerator market, NVIDIA's products offer customers the lowest Total Cost of Ownership (TCO). Kress pointed out at the JPMorgan event that enterprise data processing demands (not just AI) are also driving the need for next-generation computing technologies, which will help total investments in the tens of trillions of dollars by the end of this decade. However, these statements did not impress investors. NVIDIA's stock price fell by less than 1% at the close of the US stock market, to $187.28. The stock has risen by 39% last year. Despite optimistic prospects, there are risks in the high valuations of investments in the AI field. Institutions such as Goldman Sachs and Bank of America predict that investors may shift from overconcentrated AI leaders towards broader cyclical stocks in 2026, expanding the market landscape. A key question is whether NVIDIA can regain its foothold in the booming AI chip market in China. Export restrictions imposed by the US have severely hindered its business in China, but the Trump administration has indicated that NVIDIA can now begin selling its H200 chip to Chinese customers. Huang said, "Demand from customers is high, very high. We have started the supply chain, and the H200 is continuously being supplied on the production line." Huang stated that the details of the US government's approval are being finalized. He said that the final approval from the Chinese government will be reflected in the form of allowing Chinese companies to submit orders. The day before, NVIDIA announced detailed information about the new chips that will be launched in the second half of this year. Huang also emphasized the energy efficiency of these processors, which can provide higher performance. Meanwhile, his comments on the cooling system caused a stir on Wall Street. Huang said that the new generation AI chip Rubin's rack can be cooled purely by liquid cooling without relying on a water chiller. On Tuesday, cooling technology companies' stocks, including Trane and Johnson Controls, plummeted. Energy consumption has been another concern in the AI industry. Huang refuted concerns that data centers would result in power shortages, saying that power shortages are a normal result of the "industrial revolution" and called for increased investment in various new forms of power generation.