The year 2025 has already started, and after a week of visiting clients in the Pearl River Delta, I always wanted to write something. If you ask me what my biggest feeling is, my feeling is that the market is very vibrant, but this round of chip entrepreneurship is coming to an end.

Most of the customers we visited attended the 2025 CES. Through exchanges with customers and industry information, they all believed that AI has arrived, and the widespread application of AI technology is a highlight of this exhibition and a future trend.

AI not only occupies a core position in hardware products, but also demonstrates its strong application potential in many fields. In particular, the rise of Chinese brands is also a major highlight of this exhibition. The number of Chinese exhibitors increased significantly, reaching more than 1,300, a record high, accounting for nearly 30%.

Innovative products

The intensive launch of innovative products is also a major feature of this exhibition. The exhibition covers multiple fields such as artificial intelligence, digital health, automotive technology, and lifestyle technology, and showcases the latest achievements in scientific and technological innovation. In particular, the application of AI technology, from smart homes to AI games, from personal AI processing terminals to AI-enabled smart hardware, has demonstrated the wide application and huge potential of AI technology in various fields. The 2025 CES exhibition not only showcases the latest progress and applications of AI technology, but also highlights the rise of Chinese brands on the global technology stage.

The widespread application of AI technology has given Chinese companies a new opportunity. AI is not just about advanced artificial intelligence, nor is it limited to NVIDIA and advanced processes. AI technology should have a broader concept, and all companies and all chip processes can participate in it. AI technology is not to replace people, but to make people's lives smoother and better.

Domestic chips

Many domestic chips have been involved in the application of AI technology, including Sanwuwei's chips, and have produced valuable economic benefits.

Although the market is still vibrant and has huge market potential in the future, this round of domestic chip replacement startups is coming to an end. In the communication with customers and chip entrepreneurs, everyone agreed that the opportunity for domestic chip replacement startups has ended.

The remaining unfinished domestic high-end chips cannot be solved in a short period of time by start-up chip companies. Start-up chip companies are at a disadvantage in both manpower and funds. In comparison, listed companies or profitable chip companies have more capabilities and advantages. The low-end and mid-end chips have completed domestic substitution, and will only become more and more competitive until the reshuffle is over.

This market situation is not only seen by chip companies, but also by customers. During the visit, customers expressed their views and ideas, believing that many chip startups can only support two years, and will not introduce new chip manufacturers within two years. They will wait until the mergers and acquisitions and reshuffles of chip companies in these two years are over before introducing evaluation.

In the communication with chip entrepreneurs, they also agreed that the domestic chip replacement startup has ended, and the next step is to replace the domestic startup stage. Domestic substitution is an opportunity, and domestic substitution is a fierce battle.

I can deeply feel the anxiety and confusion of that entrepreneur. He can no longer go down the path of chip entrepreneurship. He has no money and no people. How can he compete with the existing leading domestic chip companies? He smiled and said that he could still hold on for another year. His entrepreneurial friends around him envied him and said that many chip startups could not last more than half a year.

Our generation of chip startups was born for domestic substitution. The substitution eventually turned into homogeneous involution, and the result of involution was elimination and elimination.

In the domestic substitution stage, it is effective for chip startups to rely on capital competition. As long as the chip can be made and the performance can be satisfied, it can win through price wars with the power of capital. Foreign chip companies will not do it if the gross profit is lower than a certain level, and will not continue to lose money in price wars. After entering the stage of domestic substitution, the original way of playing is simply not feasible. Unless the leading domestic chip companies give up voluntarily, they will fight the price war to the end. Later domestic chip startups are likely to be cannon fodder, and the investors' money will also be buried with them.

Therefore, if domestic chip startups cannot be the best in the subdivided track or more subdivided track, they will not have a chance. If the company cannot solve the bottleneck problem and cannot achieve profitability, the chip company can only be eliminated.

For this year's chip startups, they must either be listed or solve the bottleneck problem. If not, there is only one way left, which is to achieve sustainable development of company profitability. If you can't find a way to make a profit, general-scale financing will not solve the problem, but only delay death.

The only way for most chip startups

I see that many domestic chip startups are still on the wrong path. If they can't make a profit in the middle and low end, they will go to the high end. If they can't make a profit in the high end, they will go to the middle and low end. At this stage, there are a lot of domestic alternative chip startups everywhere, it's difficult to go anywhere. When the comic IC asked me to review the summary of 2024, I said this: Looking back on 2024, I feel that the biggest challenge still comes from the market, and the bottomless internal volume has caused the market to get out of control. Escaping from the internal volume may be the best choice, going to places where others don't go and making chips that others are unwilling to make.

We were wrong, and the investors were wrong. They all wanted to be big and complete, not small and beautiful. If each of them focused on their own segments at the beginning, and finally came to an integration and merger, each made a profit, and everyone was happy. But now all that is left is a life-and-death struggle.

Profitability is the only way for most chip startups to survive this year. As for how to make a profit, I would like to share with you the following words from Hong Jing, the founding partner of Gaocheng Investment:

“The premise for To B companies to make a profit must be focus. Focus, focus, and focus again. Become a big fish in a small pond first. Only with a base can you achieve the final victory.

In addition, do not regard a one-time upgrade of technology as a core barrier. Many people do not distinguish between entry barriers and incumbent advantages. For example, the entry barrier is that AI is very powerful. Only a returnee doctor can start an AI business. The entry barrier is high. Although the technology has advantages over the previous generation of technology, there is no essential difference between the technology entrepreneurs of that generation. At that time, the small leaders did not actually have much incumbent advantage.

So everyone looks back, the final beneficiaries of that wave of AI were Hikvision and Huawei, because Hikvision and Huawei only need 6 months or 12 months to catch up with your algorithm, but the four little dragons need 20 years to catch up. These two companies have deepened their understanding of clients, the construction of sales channels, and the support of ecological partners in the industry. Especially with the arrival of each wave of technology cycles, To In B, different technology cycles solve the same business problems, so enterprises have compound interest, and entrepreneurs can have compound interest by digging deep into the industry.

However, the arrival of each wave of technology cycles will sometimes be subverted. The subversion is because people with new technologies can acquire customers faster, or people with customers can develop new technologies faster. There is not necessarily a unified answer. It also depends on the characteristics of the industry and the speed at which the company embraces new products and new customers.

Therefore, if you want to make a profit, you must focus. Only by focusing can you continuously form your own incumbent advantage in products, marketing, and sales. Even if it is very small, even if it is very detailed, as long as you have a 40% market share, you can usually make a profit relatively quickly. Moreover, for enterprises, only by establishing a base can you obtain market share and profit pools, and can you deliver talents to other battlefields.

In short, the iron wall of To B is not built all at once, nor is it built by running blindly, but it is built by hammering down one nail at a time. "

The wheel of history rolls forward, and the trend of the times is mighty. Only by complying can you gain a foothold. To reach the other side of our ideal, we must keep moving forward along the right path. This round of chip entrepreneurship is coming to an end, but it is giving birth to new chip entrepreneurship opportunities in the future.

Source: Content transferred from the official account [Zhong Lin Talks about Chips], author: Zhong Lin

View more at EASELINK