We know that technology is the future of mankind.
In the past 40 years, there has been a major technological change almost every 10 years, and every technological wave has brought new winners, and some people have become rich as a result. For example, in the personal computer era, IBM and Microsoft became popular, and the mobile Internet era became Apple and Google. And we are now at the crossroads of technological change. As investors, whoever can grasp the next technological change will be more likely to be the winner in the next 10 years.
So what will the upcoming technological change be? I think it will be artificial intelligence AI. This is not what I said by myself. Catherine Wood, president of Ark Fund, believes that AI will be an industry in the future. Various research reports by investment banks Goldman Sachs and Morgan Stanley have also regarded AI as the most important investment layout in the next 10 years. So here comes the problem. Who will stand out from the AI technology wave in the next 10 years?
Before answering this question, let me tell you a short story. In the past, a gold mine was discovered in the central part of the United States. Many gold prospectors came here to find a gold mine. However, the people who rushed to mine basically didn't get any gold. On the contrary, the people who sold the gold shovel by the gold mine finally made money. The current AI industry is very much like a gold mine. All companies want to come in, but the problem is that their uncertainty is too high. Like the Baidu we talked about in the previous issue, it is representative of high ceilings but strong uncertainty. The company in the AI industry that we are going to talk about today is just the opposite of Baidu. He can be said to be the strongest gold shovel in the AI gold mine. He may not have such an exciting upper limit, but he is extremely certain. If you want to layout the AI industry, he is definitely a company that cannot be missed.
This is Nvidia, stock code NVDA.
Because Nvidia’s stock price has recently been overbought severely, if you only focus on short-term K-line, then this stock is definitely not for you. But if you want to understand NVIDIA's investment logic, and use this to understand the future opportunities of the AI industry. Then I promise you will gain something after reading this article.
NVIDIA and artificial intelligence
Some people may not understand, how can Nvidia, a graphics card company, become an AI company? To answer this question, we have to start with the history of Nvidia. Six years ago, Nvidia’s focus was mainly on providing high-performance graphics cards, or GPUs, to the gaming industry. At that time, Nvidia could only be regarded as a screw in the chip design industry, and no one took him seriously. But later, when AI technology appeared, people found that GPUs are more suitable for AI calculations than traditional CPUs. Because of the neural network and deep learning required by AI, a large number of matrix operations are required, and this is the strength of the GPU. The wave of technological change hits Nvidia, and opportunities are reserved for those who are prepared. Nvidia executives keenly sensed the potential for mutual cooperation between AI and GPU and decisively turned the company's strategy into a comprehensive layout in the AI industry. To this day, Nvidia can be said to be the well-deserved overlord of the entire AI industry infrastructure.
Nvidia's dominance can be discussed in terms of hardware and software. From the hardware point of view, GPU will be the future of increased computing power. It can be seen that the development of traditional CPUs has reached a bottleneck due to physical limitations, and performance can only increase by 10% per year. On the basis that GPU has surpassed CPU in current computing power, there is still high room for improvement in the future. It is expected that in 2025, it will be increased to 1,000 times the current level. As a technology that strongly relies on computing power, AI will only become more and more dependent on GPUs. And Nvidia will undoubtedly be the biggest beneficiary.
Nvidia's GPU accounts for 80% of the entire market share, and Nvidia also accounts for 70% of the world's top 500 supercomputers. The only competitor, AMD, is not only far lower than Nvidia in terms of market share, but the gap in technology is even greater, especially for GPUs used in AI. Nvidia occupies nearly 100% of the market. Although there are still a few sporadic competitors in the AI chip field, such as the TPU developed by Google specifically for AI, and an upgraded version of Apple's possible future M1 chip, Nvidia's leading position is very solid. In the foreseeable future, as long as AI still needs GPU to process data, NVIDIA will be the best company selling a gold shovel.
NVIDIA's layout and development
For the future of NVIDIA in the AI industry. To layout the AI industry, you need to think about two things. One is the disruptive changes brought about by AI technology itself. The other is the rapid expansion of the AI industry as a whole. The most frightening thing about AI technology is that it will naturally form a competitive environment where the strong are always strong. Although the initial development is time-consuming and labor-intensive, once completed, his competitive advantage will be beyond the reach of latecomers. Companies like Baidu and Google, the AI industry leader, are typical of this type of company. But Nvidia is different. He does not have an advantage in AI technology, so there will be no exciting explosive growth. The logic of investing in it is to value the rapid expansion of the AI industry as a whole and to achieve growth by selling gold mining shovels with stronger certainty. This growth rate will be considerable, and the upper limit is also very high. If you are after certainty, then Nvidia will be a better AI layout.
Nvidia's certainty is not only reflected in its AI layout. He also used the powerful computing power of his GPU to deploy many other cutting-edge technologies to make his business more decentralized. Once the gold-digging shovel is hard to sell, he still has a silver digging shovel to dig a copper shovel. Since we are analyzing the company Nvidia, we cannot do without those, except for AI, Nvidia's other layouts.
First of all, we must be clear that the traditional game business will still be the main source of Nvidia's revenue for a long time to come. The game industry currently accounts for 47% of Nvidia's total revenue, and Nvidia also relies heavily on the development of this part of the business. And the game industry is not stagnant water, his development speed is still very fast. The excitement of future development lies in the popularization of VR and AR games. Moreover, the update of game GPUs is very fast, and product performance needs to be improved every two or three years, which also leads to the continuous increase in demand for game GPUs.
Gaming is the traditional business of NVIDIA, and it is the guarantee of NVIDIA's fundamentals. But what is even more exciting is the three emerging industries that Nvidia is deploying. Each of these industries may be a trillion-level market in the next 10 years and 20 years, which is no less than its AI deployment potential.
The first industry is cloud computing.
Just as AI companies cannot do without the GPU provided by Nvidia, the cloud computing field also relies heavily on Nvidia's hardware support. In 2019, 97.4% of AI accelerators in the world's top four cloud providers AWS, Azure, Google Cloud, and Alibaba Cloud used Nvidia GPUs. The cloud computing industry still has high room for growth, and the demand for NVIDIA GPUs will only increase.
The second industry is a digital currency.
Since Nvidia GPU provides powerful matrix computing power, it is very suitable for Ethereum mining. And it is also difficult to find alternatives. Nvidia also designed a dedicated GPU specifically for Ethereum miners. I believe I don’t need to say more about the development potential of digital currency. Some analysts predict that mining GPUs will be a $5 billion market in 4 years, which is equivalent to 1/3 of Nvidia’s current revenue. The future potential cannot be underestimated.
The third industry is autonomous driving.
Autonomous driving is currently the smallest part of Nvidia's revenue, but it is also the most potential part. At present, NVIDIA cooperates with many automakers in autonomous driving, including well-known manufacturers such as Mercedes-Benz, Hyundai, Volvo, and NIO. They are building an open-source platform to provide the automotive industry with an infrastructure that can develop autonomous driving on its own. In the future, they will also develop their own autonomous driving systems. This part of the business has high potential but strong uncertainty. In my opinion, it is feasible to build an open-source autonomous driving research and development platform, but I am skeptical of developing autonomous driving technology.
It can be seen that NVIDIA is relying on its own hardware advantages to actively deploy cutting-edge technologies with the most development potential. It is not important to him whether to dig gold or not, so he concentrates on doing a good job in the business of digging gold shovels. And it seems that Nvidia's layout effect in these cutting-edge technology industries is pretty good. The financial report shows that Nvidia's data center business, that is, the layout of AI and cloud computing, has grown at an average annual rate of 80% in the past five years. And this speed will continue to increase because of Nvidia's ever-increasing competitive advantage.
NVIDIA's risk
Of course, there are not only opportunities for any company. Nvidia’s investment risk is also something we must understand. In the short term, Nvidia will face two major risks. If anything goes wrong, Nvidia's stock price will fluctuate.
The first is the acquisition of ARM mentioned above. This acquisition is very important to Nvidia, but it is still a big question mark that it can finally be realized. The ARM company involves many interest groups, and the chip is an extremely sensitive industry in the global political game. If an ARM is bought by an American company and is to be built into a super chip giant, then the interests of technology giants including the United Kingdom, the European Union, China, and the United States may all be harmed. Therefore, if any party intervenes, the acquisition cannot be completed.
The second risk is the fluctuation of digital currency prices. Now NVIDIA has made a deep layout in the digital currency industry, and its stock price is also strongly correlated with the price of digital currency. Once the price of digital currency falls sharply, then Nvidia's stock price will inevitably be affected.
In the long run, Nvidia also faces two major risks.
The first is potential competition. Although at present, Nvidia is the absolute leader in computing power. But with the rapid development of science and technology, no one can guarantee that NVIDIA can maintain such a monopoly for a long time. For example, currently, in the AI field, there is a TPU specially developed by Google for AI to threaten Nvidia's position. In the field of autonomous driving, Intel is also developing its own chip Mobileye. In the field of cloud computing, Amazon has also launched its self-developed chip Graviton. Major giants are seizing this market one after another, and it is not easy for Nvidia to maintain its monopoly position.
Second, as far as the AI industry is concerned, how fast it is developing is still unknown at present. The AI industry has a high ceiling, but it is still restricted by many factors such as funding, training time, computing power, and commercialization, and it has not been able to release its potential. Although Nvidia has a high degree of certainty in selling gold mining shovels, there is still uncertainty when it will start selling on a large scale.
NVIDIA's financial performance
After talking about Nvidia's business model, let's analyze the financial report. From the financial report, we can see that it has the characteristics of strong profitability and rapid growth. Since it began to deploy AI and other emerging technology fields in 2016, NVIDIA's revenue has grown at an average annual rate of nearly 30%, and its annualized profit growth rate has reached as high as 63%. The most surprising is the growth of its gross profit margin. In the past 12 years, Nvidia's gross profit margin has nearly doubled, from 34.3% in 2009 to 62.3% in 21, which is not easy for a hardware design company.
Nvidia can achieve this growth rate mainly for two reasons. One is because GPUs have continued to be in short supply in recent years. After Nvidia has higher bargaining power, profits have risen. The second is derived from Nvidia's data center business, which is the layout of AI and cloud computing. The gross profit margin of this business is extremely high, and in recent years, it has grown at an average annual rate of 80%. While the proportion is getting higher and higher, it has also accelerated the improvement of Nvidia's overall profit margin.
Another noteworthy point of Nvidia’s financial report is its R&D expenditures. Nvidia has today's position, and its technological advantages are indispensable. To maintain such advantages, continuous R&D investment is essential. As can be seen from the figure below, Nvidia’s R&D expenditures account for more than 20% of total revenues throughout the year. What's even rarer is that with the increase in revenue in the past two years, this ratio has risen instead of falling. In horizontal comparison, Nvidia's R&D expenditure ratio far exceeds that of companies like Tesla, Google, Baidu, and Intel that also rely on technological advantages.
At the same time, it also analyzed the financial data of Nvidia's earnings quality, repayment ability, and other financial data, and the results were very satisfactory. No hidden dangers were found in the financial report.
NVIDIA's valuation
Finally, let's take a look at Nvidia's valuation. For companies like this, with deep moats, high-speed growth, and strong certainty, they usually have one common feature, which is that is, expensive. Nvidia is no exception. Such a good company will naturally be sought after by capital. NVIDIA’s PS is 26 times, PE is 94 times, and PCF is 74 times. In contrast, Intel's PE is only 10 times, and AMD's PE is only 40 times. From the valuation point of view, Nvidia is not cheap. Although NVIDIA's growth rate and certainty far exceed its competitors. This may be related to the recent overbought of Nvidia’s stock price, or it may be due to investors’ increased expectations for Nvidia’s future growth. It can be seen that Nvidia's Forward PE will drop to 46 times in a year, which is still a very reasonable valuation range.
Finally, my opinion on investing in Nvidia.
As the most certain company in the AI industry, I think Nvidia is a company that must be deployed. However, when the layout is the key. Recently, Nvidia’s share price has seen a strong increase. It is indeed very overbought and its valuation has also been overvalued. Let's see if long-term interest rates rise in the second half of this year, or after the Fed cuts QE, is there any chance? If you are worried about going short, you can consider buying small positions in batches. After all, investing in Nvidia is to agree with its layout in cutting-edge technology, and these benefits must be the most significant in the long term. Therefore, the difference between the current stock price of 10 yuan and 20 yuan is not a big deal in the long run.
Comments
Post a Comment