The most important branch on earthstock——Nvidia.
Imagine,If you had $1,000 ten years ago, what would you choose to invest in?
The veteran media CNBC gaveup to dateThe answer is to buy Nvidia stock. As AJ Bell, investment director, said, "The people who made the most money in the gold rush in the mid-nineteenth century were those who provided the tools, not those who found the gold mines. Today, Nvidia actually plays the same role in this technological revolution."
This Thursday, Nvidia's market value soared by $277 billion in a single day, setting a record for a single day in Wall Street history.maximumThe increase easily surpassed the $196 billion that Meta Platforms had just achieved. This single-day increase also exceeded the market value of Coca-Cola's $263.7 billion, which is equivalent to the market value of Alibaba and Baidu in China.
What else can stop Nvidia?
Nvidia is unstoppable
One year of AI development is equivalent to ten years. If we want to illustrate this vividly, then Nvidia is probablymostOne of the manifestations.
In the past few years, Nvidia's stock price has been like a science fiction movie, full of surprises and excitement. If you bought $1,000 of Nvidia's stock ten years ago, based on the closing price of $694 on February 22, your return on investment would have reached $223,40%, and the price after the market on the 22nd has exceeded $800, and it may rise again.
If this is not intuitive enough, let me tell you a fact: in just two months into 2024, Nvidia's stock price has risen by 63%, and last year, Nvidia's stock price has more than tripled, and its current market value has exceeded $1.9 trillion. The reason why Nvidia's stock price is so highStrongThis may be related to the fact that in the past four quarters, Nvidia’s revenue announcements have exceeded Wall Street’s estimates by 10%-20% each time.
Unlike Sora, which has not been released yet but has already exploded on social media, NVIDIA is more like a "silently making a fortune" character, and its revenue performance comes from the GPUs sold one by one. And it is conceivable that with the subsequent launch of Sora, it will continue to widen the supply and demand gap of NVIDIA GPUs like ChatGPT.
Nvidia's stock price has not yet reached its peak.
Nvidia is surrounded by wolves
Obviously, Nvidia is the only company that controls the foundation of the AI revolution.It always makes companies that develop and apply applications feel a little insecure.
current,Nvidia's original customers are increasingly becoming competitors.Major customers such as Microsoft, Google, Meta, Amazon and Tesla are investing a lot of resources in developing their own AI chips. Even the hottest AI supernova at the moment, OpenAI, is preparing a chip project.
Altman plans to recruit talents, design chips, and even build wafer fabs from scratch. The $7 trillion financing plan he announced surprised the world. He said, "The world needs more AI computing, which requires the power of semiconductors to change... The investment is more than anyone knows or thinks, and the total cost of infrastructure will be higher than we think."
Not only did Nvidia's original customers want to get rid of it, but Nvidia was also "stolen" by its own people.A Silicon Valley startup called Grop has launched a new AI chip, LPU, claiming to have achieved "surfaceStrongestInference". The inference speed of running large models on Grop is 10 times faster than that of Nvidia's GPU, and the cost is only one-tenth of the latter. Therefore, during the same period of its release, Nvidia's stock price coincidentally fell for four consecutive trading days.
Grop's entrepreneurial team are all former employees of Nvidia. It is reported that they believe they were treated unfairly at Nvidia and have differences in technical concepts.
Recently, the revived SoftBank is planning to fight in the semiconductor industry again, targeting Nvidia and setting its sights on the hottest AI chip industry.It can be seen that the AI chip market will be even hotter in the future.
Nvidia is hard to shake
However,Although the number of competitors is gradually increasing, Nvidia's market position will be difficult to shake in the short term.
Nvidia's current net profit margin is as high as 56%. If it wants to engage in a price war, there is still a lot of room for improvement. In addition, Nvidia itself is constantly iterating and updating its technology to continue to consolidate its dominant position in AI hardware.
at the same time,Nvidia is also making multiple bets, and deepened cooperation with giants such as Google, Microsoft, and OpenAI to jointly promote the prosperity of the generative AI industry. For example, we have joined hands with Google to optimize the running performance of the open model Gemma on devices such as laptops, allowing developers to run various AI training tools in NVIDIA's toolbox with a lower threshold, and thus build their own large model products on this basis.
In addition, NVIDIA has also launched its own production AI products and has begun to provide cloud services, allowing companies to use NVIDIA's AI technology directly through cloud services.
In addition to the explosive growth in demand for generative AI, NVIDIA is also addressing more vertical needs in some more vertical industries, such as automotive, healthcare, and financial services, as well as autonomous driving, drug development, and more.There are also corresponding hardware and software platforms.
For example, China's new energy vehicle brands such as Xiaopeng and Weilai have reached cooperation with Nvidia to use the automotive chips provided by the latter. New energy vehicles are also a rapidly growing market at present. Last year, the automotive industry contributed more than US$1 billion in revenue to Nvidia's data center business.
Moreover, since NVIDIA has bound millions of AI developers and formed a powerful software ecosystem based on the computing platform CUDA, which it began developing more than ten years ago, there will be certain resistance in getting these developers to abandon NVIDIA and become familiar with other computing platforms.
therefore,As the demand for AI computing power grows, NVIDIA will be the only one that can handle most of it in the short term..