Text by Richard Nieva
Yesterday evening, Beijing time, Nvidia's stock price went crazy again, and drove U.S. technology stocks and various indexes to start a full-scale carnival; Correspondingly, "Flying Nvidia" has become the most anxious discussion among investors about the U.S. market in the past 12 hours.
In the long-short debate, there are two people who have become the focus of attention, and what they have in common is that neither of them bets on Nvidia, and they are both big enough.
According to the disclosure of position documents, the account owner "Yongping Duan" holds a total of 9** as of the end of 2023, including Apple, Berkshire-B, Google-C, Alibaba, Disney, Occidental Petroleum and other companies, among them, the position of Alibaba in the fourth quarter increased by 63%, and the amount is expected to reach 1$5.7 billion, and there is no Nvidia in other holdings.
Another focus is on Warren Buffett himself, last night Nvidia's market capitalization surpassed Berkshire Hathaway, and successfully surpassed Google and Amazon to become the third largest public company on the planet for the first time; But Warren Buffett has not been a significant investor in Nvidia, and even warned Nvidia in August last year that its market value was too high. In a way, Buffett has also stepped on Nvidia.
The shortfall of the two "conscientious" investors in the market may explain the irrational prosperity of AI-related companies in the United States, but the core question is still about how much money Nvidia has to "gamble" on.
Nvidia, in addition to its solid performance, is becoming a venture capital complex. The trillion-dollar chipmaker has closed more than 20 VC deals through its investment arm, NConventures, and its corporate development team. Companies backed by Nvidia say the deal gives them intangible assets, such as access to Nvidia's research team and even its prestigious CEO.
Nvidia posted its third straight quarter of record profits and sales in its fourth-quarter 2023 earnings report, and its shares surged to 772$65, more than 14% from Wednesday**.
Shares of Nvidia rival AMD were up 6 percent in early trading2%, while the share price of Arm, whose chip architecture powers Apple's chips, Nvidia's Grace CPUs and almost all smartphones, is up nearly 9%. Shares of other Silicon Valley giants such as Intel and Qualcomm rose to a lesser level, at 27% and 17%。
Considering Nvidia's position among the world's "most important**", its boom is seen as a major catalyst for broader market sentiment, and it has an important boost not only for other tech companies, but for all**.
01.Nvidia became the engine of the U.S. stocks
Last night, the benchmark S&P 500** exceeded 2%, hitting an all-time high of nearly 5,100 points; The tech-heavy Nasdaq jumped more than 2%; Even the Dow Jones Industrial **erage, which excludes Nvidia, climbed more than 1% to a record high of more than 39,000 points.
TSMC, the main manufacturer of Nvidia's chips, went public on the New York Stock Exchange on Thursday 376%, while Asml Holdings, a Dutch manufacturer of advanced chip manufacturing equipment, traded 37%。
In the quarter that ended last month, Nvidia achieved $22 billion in sales and $13 billion in net profit, with annual revenue up 265% and annual profit up 769%, easily beating analysts' consensus expectations. Its growth was largely driven by its booming data center division, which includes Nvidia's popular graphics processing units (GPUs), which power generative AI, and its revenue grew 409% annually in the last quarter to $18.4 billion. The company's financial numbers are impressive, following growing investor and corporate interest in artificial intelligence that has seen its share price at historic speeds**, while stock indices are also strong**.
Nvidia's record profits and growth stemmed from its custom-designed chips being used to train and support AI services operated by companies such as OpenAI, Google, Microsoft, and Amazon. However, Nvidia's near-dominant position and high demand have raised some concerns as AI companies have had to deal with chip shortages. This has prompted companies like Microsoft and Meta to look to the likes of AMD for alternatives. Last year, the two companies, along with OpenAI, announced plans to use AMD's Instinct Mi300X chip as an alternative to Nvidia's popular H100 chip.
Jensen Huang, co-founder and CEO of NVIDIA, at NVIDIA's headquarters in Santa Carla, California. **ethan pines for forbes
Bernstein analyst Stacy Rasgon reviewed Nvidia's "extraordinary" first-quarter results in a note to clients on Thursday, adding that he believes the company's growth prospects remain "solid."
02.How NVIDIA's investment arm is driving a new batch of AI startups
When AI biotech company Terray Therapeutics first trained its AI model, Coati, early last year, CEO Jacob Berlin dismissed the technology as "not bad." It's practical, but it could be done better.
But that was before Nvidia, a trillion-dollar chipmaker, invested in the startup, which focuses on small molecule drug development. After announcing the investment in November last year, Terray took advantage of the enhanced computing resources and NVIDIA's engineering prowess to completely retrain the model. In an interview with Forbes, Berlin said: "We're seeing better performance [after adopting Nvidia's technology]. We wouldn't have been able to achieve this without our partnership with NVIDIA and their support. ”
The revamp of this AI model illustrates that there are many intangible benefits to getting investment from the world's most valuable chipmakers beyond funding. Over the past two years, Nvidia has invested in more than two dozen companies through its venture capital arm, nVentures, and its broader enterprise development team. This is the case for these companies.
In Terray's case, the most important factor was access to Nvidia's researchers and engineers, who knew how to get the most out of the chip giant's supplied-in GPUs (graphics processing units, which can be used for large-scale AI development). At some point, that also means getting a quick email response from Nvidia CEO Jensen Huang. The CEO-led committee approves Nvidia's investment, and his influence in the tech industry is growing by the day. For many startups, this means gaining recognition from one of the most important companies in the AI boom.
In an interview with Forbes, Sid Siddeek, vice president and head of NConventures at Nvidia, said: "Our investments are broad – 'broad' in the broadest sense of the word. We don't just give money, we give expertise, technology, benefits, and all sorts of things. One goal of these investments, he noted, is to "generate healthy returns for Nvidia, just like any financial investor." ”。
Many big tech companies have their own venture capital divisions, including Alphabet, Microsoft and Salesforce. But some of the investee companies say Nvidia is relatively new in the space and can offer benefits over traditional VC firms in some ways. NVentures was founded in early 2022, and a few months later, OpenAI's ChatGPT (which is mostly trained on Nvidia's GPUs) was released, setting off a wave in Silicon Valley that brought AI into the mainstream. Since then, Nvidia has been investing heavily, including top AI unicorns Cohere, Hugging Face and Inflection.
"At that time, there was a little thing called generative AI that was starting to catch on," Siddeek said. It can start from scratch and grow indefinitely. In December, the company said it had announced 14 investments since the beginning of 2023, in areas such as health, enterprise and logistics. Meanwhile, NCoptive's ** only opened last month, which shows that everything about it is still in its infancy.
Siddeek told Forbes that Nventures had "made several investments" in the new year, but he declined to provide further details.
Forbes interviewed seven companies that have been invested by Nvidia, and they declined to disclose the specific terms of the deal with the chip giant. Some companies claim to have been given more access to Nvidia chips, which are currently the most sought-after resources in Silicon Valley, but they did not disclose specific details. Other companies said they were not given special treatment.
Jorge Torres, CEO of startup MindsDB, said: "It may be an exaggeration to say it's a 'VIP pass', but it's better to have a relationship than if you don't. "MindsDB is a company that allows developers to connect AI models with data sources.
Nventures told Forbes that they are not giving priority to portfolio companies to use GPUs. As startups and tech giants wage an arms race to train the most powerful AI models, GPUs have become a new form of currency. But Kanjun Qiu, CEO of Imbue AI, revealed that this does not mean that the companies invested by Nvidia did not make such inquiries during the deal. IMBUE AI is a startup that aims to develop artificial intelligence capable of reasoning like humans. "We now know that we won't get priority access. But we've also asked, "....Will we have priority access? Qiu said with a smile. Nvidia and the founders of several portfolio companies said that Nvidia did not require or designate startups to spend money on Nvidia chips or other products.
The funds used by nVentures and the corporate development team to invest come from NVIDIA's balance sheet. Now is a good time to do this. According to research firm Omdia, Nvidia is the global leader in GPUs, accounting for more than 70% of AI chip sales. In the first three quarters of last year, the company's revenue reached 388$200 million (Nvidia will report fourth-quarter earnings at the end of the month). The company's sales in the third quarter were up more than 200% from the year-ago quarter. The result was a significant rise in market capitalization, soaring to 1$8 trillion, making it the fourth-highest company in the world by market capitalization, surpassing Alphabet, Amazon, and Meta.
03.What role does Huang play in investing?
There is another attraction of being invested by Nvidia: Nvidia leader Jensen Huang, who loves to wear leather jackets. When it comes to reaching out to the CEO, the experience of different investee companies varies. For some portfolio companies, such as Imbue and Utilidata, a company that aims to reshape the energy grid with artificial intelligence, Huang was personally involved in the investment process and met with the founders. Mr. Qiu, the chief executive of Imbue, said she has been in contact with Mr. Huang ever since. On one occasion, she emailed Mr. Huang asking how much the company should pay executives, and Mr. Huang was quick to respond. "He was very accommodating. I tried not to bother him, but I was sure to be able to get in touch with him. ”
On the other hand, the founders of other portfolio companies say they have never met him. "You have to remember that Huang is a busy man. Siddeek said. "I'm talking about the obvious. However, he added, Huang's investment committee is responsible for signing every deal. Asked about Huang's criteria for meeting with the founders during the deal, Nvidia declined to comment.
In addition to Huang, some of the founders of portfolio companies said they were able to get a quick response from Nvidia because they were invested in them. "It's clear that within Nvidia, people are very responsive to requests for help from the nventures team. Bob Hall, CTO of Outrider, a company specializing in autonomous freight and logistics, said. "So their involvement, and the importance of nventures within NVIDIA, has been a huge boost for us. ”
Nvidia may be the tech industry leader right now, but the new world order is common in Silicon Valley (just ask Yahoo, MySpace or Amd, Huang's former employer). As other companies try to develop viable alternatives to Nvidia's chips, the company may see investing in startups as a way to diversify its business in case Nvidia's GPUs no longer dominate the market as much as they do today. Three startup founders who have been backed by nVentures told Forbes that they think that could be the case.
A founder, who asked for anonymity to speak candidly about Nvidia, said:"Nvidia will be very difficult to replace in the coming generations, but there will certainly be some very smart people who will try to do it. ”
Siddeek admits that from an investment perspective, Nvidia is looking to the future. He said:"We're looking for founders who really have a bold vision, who don't see the world today, but the world in 5 or 10 years. ”
Nvidia has a long history in this regard. The company was founded 30 years ago with the initial goal of utilizing its processors to make the graphics of ** games better. In recent years, GPUs have begun to be used for cryptocurrency mining, and Nvidia's business has leaned in this direction. Of course, GPUs have recently become the driving engine for AI training.
The discovery first appeared in 2009, when a group of Stanford professors, including Andrew Ng, realized that the efficient computational structure of GPUs could speed up the training of machine learning. Ng later co-founded the Google Brain Research Lab. For Nvidia, this was partly a result of the right time, but the company's success is also due to Huang taking the right steps to seize the opportunity of the coming AI revolution.
Terray's Berlin said"It's clear that they're making the right decisions when it comes to picking verticals that they think will grow and take off in response to the demand for chips and software. They are trying to develop another multi-billion dollar vertical market. ”
Alex Konrad, Iain Martin and Kenrick Cai contributed to this article.
This article is translated from.