share_log

英伟达下轮暴涨的动力?主权AI崛起!今年将创收数十亿美元

The impetus for Nvidia's next surge? The rise of sovereign AI! Will generate billions of dollars in revenue this year

Zhitong Finance ·  May 23 12:05

This time, in addition to AI chips, Nvidia has given a new impetus for growth.

It's no exaggeration to say,$NVIDIA (NVDA.US)$Every financial report can revive investors' “faith in AI,” and this one is just as shocking! This time, in addition to AI chips, Nvidia has given a new impetus for growth: sovereign AI and new multi-modal capabilities. Against the backdrop of burgeoning demand for AI chips, Nvidia's ambition to promote business diversification through AI sovereignty has given it a bit more momentum to increase dividends and stock splits.

Chipmaker Nvidia, which is at the center of the artificial intelligence boom, once rose by about 6% after the market on Thursday, with a market capitalization increase of nearly 140 billion US dollars. This means that the after-market increase alone surpassed Intel's overall market value. What's even more surprising is that Nvidia optimistically expects sales of 28 billion US dollars in the next quarter to be more than double Intel's guidelines.

Demand for AI chips is booming. New chip platforms will generate revenue this year and eliminate the “empty window period” of demand

Before the results were announced, one of the major concerns of the market about Nvidia's earnings report was whether its latest data could justify the dizzying rise in its stock price. As of Wednesday's close, the stock has accumulated a 92% increase this year, and investors hope the company will continue to break expectations. The latest report did not disappoint, and CEO Hwang In-hoon talked about the dawning of a new era, thus arousing excitement. “This is the beginning of a new industrial revolution,” he said in an interview, which echoes one of his favorite themes; “It's really exciting.”

Results for the first fiscal quarter ended April 28 exceeded expectations, thanks to growth in Nvidia's data center division. The company said on Wednesday that second-quarter revenue will reach around $28 billion, exceeding analysts' forecasts of $26.8 billion. Moreover, optimistic sales forecasts show that the company's spending on artificial intelligence computing remains strong.

This optimistic outlook reinforces Nvidia's position as the biggest beneficiary of AI spending. The company's so-called artificial intelligence accelerators — chips that help data centers develop chatbots and other cutting-edge tools — have become a hot item over the past two years, causing sales to soar. In the first fiscal quarter, Nvidia's revenue more than tripled to $26 billion. Excluding certain items, the profit was $6.12 per share, far higher than analysts' expectations of revenue of about US$24.7 billion and earnings per share of US$5.65.

Nvidia's data center division — currently its largest source of sales — generated $22.6 billion in revenue last quarter. The gaming division provided $2.6 billion in revenue. The expectations given by analysts are $21 billion for the data center sector and $2.6 billion for the gaming sector.

The Blackwell platform achieved revenue this year, eliminating concerns about an “empty window” in demand

The company's new chip platform, called Blackwell, is now in full production. It lays the foundation for generative artificial intelligence that can handle trillions of parameters. Wong In-hoon said, “We are ready for the next wave of growth.”

Wong In-hoon pointed out that the company will receive a “significant amount” of revenue in 2024 from the new Blackwell chip, but customers are not reducing orders for existing products — something that some analysts are concerned about. These buyers need the current generation to help them build infrastructure as soon as possible, he said.

As Blackwell enters production, some investors are worried that Nvidia may face a sales “gap”: some customers will choose to wait for Blackwell and abandon Hopper. In response to this concern, Hwang In-hoon emphasized during the conference call that Blackwell can be “backwards compatible” with Hopper's architectural system. This means that customers don't need to “wait patiently” for Blackwell to ship, buy Hopper early and enjoy it early, and it will also be “silky” if they want to transition to Blackwell later.

Demand for Nvidia products has already surpassed supply, and Wong In-hoon expects this situation to continue until next year. He said the company is also facing other challenges: its technology is becoming increasingly complex and now includes complete computer systems, which means its supply chain is much more complex. This makes it more difficult to increase output. “No one has ever mass-produced supercomputing chips,” he said in an interview. We're doing what we can.”

The stock split raised interest rates, and Nvidia's stock price broke through 1,000 US dollars, leading the way for AI stocks

The company announced a 150% increase in its quarterly dividend to 10 cents per share. Nvidia also announced a 10-to-1 split of shares, which helped drive its share price higher. The company said the move was aimed at making shares more accessible to employees and investors. This will be Nvidia's second spin-off in the past three years, and since the last spin-off was announced in May 2021, the company's stock price has risen by more than 500%.

In theory, stock splits have no fundamental effect on the value of shares. However, Alec Young, chief investment strategist at Mapsignals, said they often help drive purchases from home investors. In an interview, Young said, “The Nvidia stock split is big news. You shouldn't underestimate the appeal of stocks favored by retail investors.”

Nvidia released another earnings report and forecast that surpassed expectations to dispel concerns that the boom in spending on data center equipment over the past year might slow down. Nvidia gave traders the green light. They are betting that there is room for continued gains in AI computing stocks, not to mention the company's own stock.

In after-market trading of US stocks on Wednesday, the stock once rose 7.8%, breaking the $1,000 mark for the first time. Nvidia's market capitalization also soared, surpassing $2.3 trillion. The rise in Nvidia's stock price helped push up the share prices of other artificial intelligence-related companies.$Super Micro Computer (SMCI.US)$,$Advanced Micro Devices (AMD.US)$und$Dell Technologies (DELL.US)$They all rose after Nvidia's results were announced.

After the US market on Wednesday, Nvidia's performance drove hardware stocks to rise sharply. The biggest increases were made by server makers Ultramicrocomputers and Dell Technologies, both of which rose by more than 4%. The manufacturer of Nvidia chips$Taiwan Semiconductor (TSM.US)$The increase was more than 3%, chip makers$Broadcom (AVGO.US)$,$Marvell Technology (MRVL.US)$With AMD, the increase was more than 2%.

In the early stages of diversification of Nvidia's business, leveraging sovereign AI

Nvidia has become Wall Street's most promising target for artificial intelligence. The sales and net profit generated over the past few quarters have greatly exceeded analysts' general expectations, and the performance forecasts provided for several consecutive quarters have shocked Wall Street. In addition to providing chips to hyperscale cloud computing service providers, Nvidia is now also trying to diversify its business — focusing on the “AI localization” market. This is also the first time this year that Hwang In-hoon mentioned “sovereign AI”, implying a surge in demand for national-level artificial intelligence hardware.

Bet on sovereign artificial intelligence that will contribute billions of dollars this year

OpenAI's ChatGPT chatbot was released in 2022, sparking a race among major tech companies to build their own artificial intelligence infrastructure. This battle made Nvidia's H100 chip a must-have product. Each chip costs tens of thousands of dollars and is often in short supply. Nvidia's chips are ideal for developing artificial intelligence software, which helps open up a new market and keep it one step ahead of competitors.

Currently, most of the new revenue comes from a small number of customers.$Amazon (AMZN.US)$,$Meta Platforms (META.US)$,$Microsoft (MSFT.US)$and Alphabet's$Alphabet-A (GOOGL.US)$These four companies are Nvidia's biggest buyers, accounting for around 40% of sales. However, Hwang In-hoon is trying to expand his bet by producing complete computers, software, and services to help more businesses and government agencies deploy their own artificial intelligence systems.

Nvidia emphasized on Wednesday that it wants to sell its technology to a wider market — beyond large cloud computing service providers known as “hyperscale.” However, Hwang In-hoon pointed out that artificial intelligence is shifting to consumer internet companies, car manufacturers, biotech, and healthcare customers. Countries are also developing their own systems, and this trend is known as sovereign artificial intelligence. He said these opportunities are creating multiple multi-billion dollar markets beyond cloud service providers.

Today, Nvidia is looking for new markets and large government-level customers to maintain its rapid growth momentum. For example, under Elon Musk$Tesla (TSLA.US)$The large-scale deployment of Nvidia chips is a sign of this expansion. The car manufacturer is using Nvidia equipment to develop software for autonomous cars.

Just last week, it was reported that Tesla had been developing plans to build a data center in China to train the algorithms needed to run fully automated vehicles (FSD) in the region. According to some sources, the electric vehicle giant is already in talks with Nvidia (NVDA.US) to buy graphics processing units in its Chinese data center. According to reports, Tesla has been making plans to establish a data center in China to train the algorithms required for FSD. According to reports, Tesla is still trying to obtain approval from Chinese regulators. If Tesla can find a workaround and provide a full version of the FSD in China, it can increase revenue and profit margins in China.

In the last quarter, hyperscale cloud computing service providers remained a key growth driver for Nvidia. They account for approximately 45% of the company's data center revenue. This shows that Nvidia is in the early stages of diversifying its business.

After Nvidia announced the results, Tim Ghriskey, senior portfolio strategist at Ingalls & Snyder, said: “We are in the midst of a technological revolution, and we are still in the early stages. It's really hard not to be optimistic right now, especially with Nvidia.”

Huang Renxun first proposed “sovereign AI capabilities” in 2024, implying a surge in demand for national-level artificial intelligence hardware

In February of this year, Nvidia CEO Hwang In-hoon said in an interview with the media that today, countries around the world are planning to establish and operate their own artificial intelligence infrastructure within their own countries, which will comprehensively boost demand for Nvidia hardware products. In an interview, Hwang In-hoon said that countries including India, Japan, France, and Canada are all talking about the importance of investing in “sovereign artificial intelligence capabilities.” “From their natural resources to their data, it's seen as something that should be refined and produced for their country. Recognition of sovereign AI capabilities is a global concept.”

Hwang In-hoon said that all countries now find themselves in increasing demand for supercomputers, which are necessary to utilize their academic achievements. As the co-founder of Nvidia, the world's first chip giant to break the trillion-dollar market cap, Hwang In-hoon has been talking about the need for countries and large technology companies in these countries to keep valuable data and intelligence locally at the beginning of this year. According to Hwang In-hoon, this national-level approach to the global AI technology layout boom will drive the pace of data center expansion, and this will require Nvidia's expertise and the core acceleration hardware of current data centers — AI chips.

Wong In-hoon now plans to expand Nvidia's government-level customer base by persuading large enterprises and government agencies to establish their own infrastructure. Hwang In-hoon said, “The vast majority of the computing market and large data centers are concentrated in the US. But this is the first time, because of the advent of computational technology that generates artificial intelligence, computing power demand will affect every country. As a result, some markets will be quite large and global.”

Hwang In-hoon's comments mean that demand for generative artificial intelligence computing is expanding from cloud computing providers to other industries such as automobiles and healthcare, and these industries often require “localization” of data. These results provided further evidence to Michael Sansoterra, chief investment officer of Silvant Capital Management, that the rise in artificial intelligence will continue. He said, “This not only shows that data centers are spending strongly on artificial intelligence, but also that people working on the periphery are likely to continue to do well. This bodes well for the healthy development of the artificial intelligence market, which is still in its early stages.”

edit/new

The translation is provided by third-party software.


The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.
    Write a comment