Source: FT Chinese

At an online event last week, Huang Renxun had a conversation with the CEO of DataDirect Networks. In this conversation, Huang Renxun had a conversation with the first time DeepSeek responded to the impact of Nvidia.

About a month ago, DeepSeek released an open source version of its R1 model. After that, Nvidia's stock price seemed to be on a roller coaster, and its stock price plummeted by 17%, which caused Nvidia's market value to evaporate by nearly $600 billion in three days.

About a month later, Nvidia's stock price has basically "recovered lost ground". Huang Renxun also mentioned DeepSeek for the first time. He said the R1 launch was "incredibly exciting" and said the previous market reaction was a misunderstanding. "I think the market's reaction to R1 at the time was 'Oh my god, the AI ​​is over! It fell from the sky, we no longer need to do any calculations!'" Huang Renxun said, "and the facts are exactly the opposite."

Huang Renxun believes that the launch of the DeepSeek R1 model has accelerated the popularity of artificial intelligence, which does not mean that the market no longer needs computing resources, but on the contrary, driven by DeepSeek, the market will need more in the future. computing power.

Huang Renxun came out to speak out this time and talked about the influence of DeepSeek, and chose a more clever time point. First, Nvidia has almost emerged from its stock price plummeting; second, Nvidia will release its fourth-quarter financial report in a few days. To the market, Nvidia's financial report has become as important as key economic data. Market analysts will judge the prosperity of the AI ​​market based on the data, and Nvidia's stock price will fluctuate accordingly.

But this time, market analysts seem to have a lot of differences in their prospects for Nvidia's financial report and stock price. The emergence of DeepSeek means the possibility of low-cost large-scale model training methods. This breaks the moat created by giants with a lot of capital, chips and computing power. In the training of large models, if the expected performance can be achieved with less computing power, then the market will not need so many high-end AI chips in the future. Will Nvidia's stock price squeeze out a bubble?

It seems that the rise of DeepSeek led to Nvidia's plunge is based on the more direct market feedback generated by this analysis. Is this a "misunderstanding"? It may take time to verify whether it is a "misunderstanding" .

But one month after DeepSeek suddenly became popular before the Spring Festival, everyone's judgment on the market is changing. Indeed, as Huang Renxun said, the release of DeepSeek's new open source model has set off a new wave of AI, and the scope of AI application is expanding, which means more computing resources are needed in the future.

Analysts who hold this view will use the Jevins paradox in economic theory to summarize this counterintuitive phenomenon of resource efficiency: when resource utilization efficiency increases,on the other hand may lead to an increase in resource consumption rather than a decrease in resource consumption. Computing power is also a resource. Driven by technological progress, computing power efficiency is improving and the market is further expanding, which in turn prompts the market's demand for computing power to continue to increase.

This is being gradually verified by the market. According to a report by the Economic Observer a week ago, more and more Chinese and foreign companies are connecting to DeepSeek, and the rapidly growing number of customers has also put a lot of pressure on servers on various platforms. The existing computing power can no longer keep up. With the demand of users, Nvidia's GPU is becoming in short supply again. "Even the stock banned models have been stolen."

Will the Jevins paradox fail in the field of AI in the future? This is why investors disagree on Nvidia's expectations. In the long run, whether the demand for computing power and chips will continue to increase depends on the increase in demand brought by AI application expansion and whether it can exceed the rate of decline in computing power costs. But nowadays, AI competition is included in the narrative of the game between great powers. When technological competition is mixed with factors, everything will become more complicated and has become apart from the purely market logic category.

According to previous media data, several American technology giants have reached historical peaks in the past year, and more than half of them may be "burned" on computing power. Although the market has been questioning the return on its computing power expenditure, its vigilance against AI and the rise of DeepSeek will still make giants insist on spending money to pile up computing power because they are afraid of falling behind. Google, Amazon and Meta's investment in computing power in the new year is still increasing, and the United States is also vigorously promoting Stargate, an AI infrastructure plan that claims to be as high as $500 billion.

At least, judging from the current situation, the rise of DeepSeek may not only not "kick Nvidia and slap the Star Gate in the face", but may instead give Nvidia an assist. Maybe even Trump Maybe they have to boast that they are "foresighted".