A few days ago, I shared two articles about the logic of the market trend in A-shares over the next ten years, focusing on technology competition. Not long after finishing the article, I happened to see a highly discussed live stream: a renowned finance professor had a dialogue with the well-known financial commentator Shui Pi.

Actually, there weren't many new insights in their discussion, and the current situation of many retail investors in the A-shares stock market was not surprising at all—regardless of whether the market is bearish or bullish, even if participating in the hottest technology sectors, the majority of retail investors will have the same outcome: whether in a bull or bear market, regardless of the sector, they will always incur losses, ultimately contributing whatever hard-earned savings they have managed to accumulate.

This outcome is exactly the basis of the core viewpoint I shared in the article:

Using the capital market to mobilize private capital to support national development in technology sectors is completely feasible.

However, what Shui Pi mentioned about the comparison of price-to-earnings ratios of Chinese and American technology companies piqued my curiosity, so I compared GPU companies from both sides.

Among the Chinese companies in the GPU sector, I feel that the currently strongest one (especially on the training side) and the one most compatible with domestic large models may not be those already listed but a telecommunications company based in Shenzhen. However, that company is not publicly listed, and we cannot see relevant financial data, so I do not consider it as an evaluation target.

The strength of those listed domestic companies may not be as strong as that Shenzhen company, but their significance in other aspects is much greater:

On one hand, they can reflect some important conditions of our country's GPU industry to a certain extent;

On the other hand, they can serve to support national technological competition and leverage financial markets to lift technology companies and create demonstration effects.

So here, regarding our GPU companies, I selected a few typical listed ones.

For American companies, I chose NVIDIA; for Chinese companies, I chose Cambricon, Moore Threads, and Mu Xi Co., Ltd.

As of January 25, 2026, NVIDIA's latest (2025) annual net profit is 120 billion USD, with a price-to-earnings ratio of around 30 times.

As of February 7, 2026, Cambricon's latest (2025) annual net profit is 2.1 billion RMB, with a price-to-earnings ratio of about 234 times.

Meanwhile, Moore Threads and Mu Xi Co., Ltd. are currently in a state of net loss.

No matter what the government's position is on the stock market and how it plans to use financial markets to fund technology sectors, these are all "political accounts."

But back to reality, back to the investment level, focusing on each investor's interests, we still need to calculate the economic account.

The so-called "economic account" means that after the noise settles, the capital market will still return to whether the company can be profitable and whether it can make money.

Once the economic account is calculated, the capital market's performance is quite brutal.

Look at the United States; merely having good data for one quarter is not enough—if the next quarter’s data has concerns, Wall Street will react harshly.

The same goes for A-shares, but there may be a time lag.

What kind of lag?

Wall Street may give you a quarter to prove that there’s no issue. A-shares might be “merciful,” allowing the bullets to fly for a while longer, giving you a year or even longer to prove that there's no problem.

But if after a year or two you still can't prove it, A-shares will still "kick you while you're down."

So, let’s take a look at Cambricon, which is currently already profitable in the GPU sector.

Cambricon's current profit is 2.1 billion RMB, but its current price-to-earnings ratio has reached 234 times, with a market value of about 470 billion RMB.

If we measure it by NVIDIA’s 30 times price-to-earnings ratio, to match the current market value, it needs to have a price-to-earnings ratio of 30, which means its annual net profit must reach 15.6 billion RMB.

That is more than 7 times the net profit for 2025.

So the next question is:

How long will it take Cambricon to achieve a net profit growth of 6 times from 2.1 billion to 15.6 billion RMB?

On the other hand, how long will the A-share capital market tolerate it achieving this goal?

One year, two years, or three years?

If it cannot achieve it within the tolerance period of the capital market, the brutality of the capital market will be realized.

Moreover, even if it achieves it in the future, that would merely match today's market value.

This is still about the already profitable Cambricon; what about Moore Threads and Mu Xi, which have yet to make a profit?

The challenges are even greater, and the potential challenges faced in the future will be harsher and riskier.

Therefore, the current enthusiasm for the technology sector in A-shares should translate into the interests of investors, and there are many issues worth observing and contemplating.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。