Nvidia is too explosive.
In one night, Nvidia's market value soared by $260 billion, approaching RMB 2 trillion. In just one night, Nvidia has risen to the market value of Moutai. In terms of Chinese concept stocks, NVIDIA's market value in just one night** is equivalent to the sum of the market capitalization of Alibaba and Meituan. As the head of Nvidia, Huang's worth has also soared to $69.4 billion, surpassing Nongfu Spring's boss Zhong Sui to become the richest man in the new generation of Chinese.
The total market capitalization of Nvidia has also come to 1At 96 trillion US dollars, it is the technology company with the highest market value founded by the Chinese, and it is also the highest achievement created by the Chinese among American chip companies. Nvidia almost single-handedly led the violence of the entire American technology sector**, and even Masayoshi Son, who lost money, also turned over.
Because AI exploded, Arm Company, which Son Masayoshi bet heavily on, also soared with Nvidia. Such a high-tech company cannot find a benchmark in the entire A-share and Chinese concept stocks. AI companies such as Nvidia, OpenAI, and ARM are really doing AI, while China's AI big Vs and A-share AI concept stocks are just hyping AI and cutting leeks.
Why is NVIDIA hailed as the cornerstone of the AI industry? How big is NVIDIA's contribution to the transformation of the AI industry? All of this has to start with the most inconspicuous game that is ridiculed by the majority of parents. In China, gaming is an under-regarded industry. Jack Ma once said publicly: "Ali doesn't make games." But the fact is that Ma Yun broke his promise, Ali has already made games, "Three Kingdoms Strategy" has only rushed to the game hot list many times, and Tencent Games is infamous.
But the rise of NVIDIA started with games. Huang's ancestral home is Zhejiang, but he was born in 1963 in Tainan City, Taiwan Province. His father was a chemical engineer and his mother was a teacher. In 1972, when his family was forced to move to Thailand due to his father's job transfer, Huang and his brother were sent to foster care with relatives in the United States.
Because of their poor relatives, 9-year-old Huang and his brother were sent to a rural boarding school in the United States. This school is a garbage school in the United States, the school spirit is very poor, elementary school students start smoking, drinking, perming their hair, getting tattoos, Huang Jenxun mingles with these bad children all day long, and his study life is also a mess, except for not breaking the law and committing crimes, all other bad habits are infected.
But fortunately, two years later, his father immigrated to the United States, which sent Huang to a regular school in the United States. Perhaps because of the talent of the Chinese, after arriving at the regular school, Huang Jenxun began to fight back wildly. First of all, his academic performance is very good, and he is far ahead of the curve. Second, he played particularly well in table tennis, with 15-year-old Huang finishing third in doubles at the US Open.
In the United States, the most popular student is this kind of dual study of civil and military, not only good in cultural classes, but also has sports expertise, only 16-year-old Huang was admitted to Oregon State University as a gifted teenager, majoring in electrical engineering. We were still in our first year of high school at the age of 16, and we were in college at the age of 16. As a result, 20-year-old Huang graduated with a bachelor's degree. After graduating from university, he went to Silicon Valley in the United States and joined AMD as a chip design engineer. But instead of going to work all the time, he went on to pursue a master's degree in electrical engineering at Stanford University.
In 1993, Huang was just 30 years old. In this year, he founded NVIDIA, determined to build graphics processing chips. However, at that time, Microsoft's Windows operating system had not yet risen, the personal computer market had not exploded, and no one was optimistic about NVIDIA's graphics processing chips. Huang searched Silicon Valley, and only two ** companies were willing to invest in Nvidia, and only one invested $2.2 million in the end. With this money, Nvidia developed the first NV-1 chip, but the sales were dismal and it could not be sold at all.
Because it is a GPU chip that integrates multiple functions such as 2D, 3D, sound card and controller control, etc., it has many functions, and it is too expensive. The failure of this one product almost wiped Nvidia bankrupt. $2.2 million was burned and nothing was found. At this life-and-death juncture, the Japanese game giant took a fancy to the controller control function of the NVIDIA chip, and gave Nvidia a deposit of $7 million, asking Nvidia to develop a special chip for their Saturn game console.
This money became a lifesaver for NVIDIA. However, due to a misjudgment of the technical direction, the development of the NV-2 chip failed, and $7 million was burned out. But the Japanese are very persistent and gambling, and Sega Games not only did not blame Nvidia, but the $7 million deposit was gone, and the rest of the balance was all given to Huang Jenxun and continued to support Nvidia. With the full support of Sega Games, NVIDIA developed the NV-3 chip. With this chip, Nvidia won a large number of orders for Sega games, not only survived, but also began to make money.
But it was Microsoft that really brought Nvidia to the first peak, as the windos graphical desktop operating system began to sweep the world. In 1999, NVIDIA developed the world's first GPU, which became a general-purpose GPU for personal computers and gaming computers. Windows ushered in the era of personal computers, and Nvidia followed Microsoft closely into thousands of households. The promotion of personal computers has also brought about the rise of the online game industry, and NVIDIA has become the world's largest game graphics card manufacturer.
Because the large-scale online games developed by game manufacturers have higher and higher requirements for image quality and graphics, the technical iteration of graphics cards is getting faster and faster. In order to match the image quality of the game, NVIDIA is forced to develop more powerful graphics cards, and stronger graphics cards can prompt game manufacturers to develop games with more beautiful graphics graphics and attract more players to recharge and renew.
Nvidia is a super GPU chip overlord raised by games!
So, how did NVIDIA transform from a game GPU chip overlord to the world's No. 1 AI chip manufacturer? There are two reasons: First, GPUs cost less than CPUs. In a Google experiment, it takes 160,000 CPU chips, the total cost is up to one million dollars, and to change to GPU, only 64 graphics cards are needed. This cost advantage allows GPUs to crush CPUs. Second, what GPUs are best at is large-scale, simple and repetitive training. In the AI model, in order to accurately recognize a face, a large number of simple and repetitive shape recognition is required, which is precisely the inherent advantage of GPU.
It is still Microsoft that pushes Nvidia to the second peak. It was Microsoft that spent $10 billion to smash OpenAI and buy Nvidia's GPU chips with huge sums of money, which led to the explosion of AI industrialization. Behind the GPT and SORA we see are hundreds of thousands of AI chips produced by Nvidia. Why has the AI gap between China and the United States widened in recent years? It is because the United States has blocked NVIDIA's top AI computing chips.
Previously, the A100 and H100, which had been blocked, were no longer NVIDIA's most advanced AI chips. Now, they've iterated the GH-200 chip. These chips are like the foundation, without the support of the underlying computing power, the entire AI industry cannot run.
For China, how to build a domestic version of Nvidia is more important than building a large AI model.