Is Foxconn really going to lose Nvidia after losing 600 billion yuan in market value in half a month?

On November 24th, the stock price of Industrial Fulian, a A-share listed company controlled by Foxconn Technology Group (also known as “Hon Hai Technology Group”), suffered a heavy blow, with a limit down at one point during trading and a closing drop of 7.8% as of 15:00.

It is understood that the sharp decline is due to the recent fermentation of “short essays”. Last week, there were continuous reports that Industrial Fulian would lower its fourth quarter cabinet shipments and performance outlook. Related rumors suggest that major clients of AI chip giant Nvidia and other companies will reduce or even stop purchasing industrial Fulian L6-L12 cabinets and instead produce them themselves. Therefore, the cabinet shipment quantity of Industrial Fulian in the fourth quarter will be reduced to 80000, and the performance outlook will also be lowered accordingly.

富士康

On October 29th, the company just released its third quarter report, showing that its revenue for the first three quarters was 603.931 billion yuan, a year-on-year increase of 38.4%; The net profit attributable to the parent company reached 22.487 billion yuan, a year-on-year increase of 48.52%. Among them, the single quarter revenue in the third quarter reached 243.172 billion yuan, a year-on-year increase of 42.81%; The net profit attributable to the parent company increased by 62.04% year-on-year to 10.373 billion yuan, achieving a single quarter net profit of over 10 billion yuan for the first time in history.

The company explained that the high performance was mainly stimulated by the continuous expansion of the AI server market, the large-scale delivery of new generation AI cabinet products for ultra large scale data centers, and strong demand for AI computing power. According to the financial report, the revenue of cloud computing business in the third quarter increased by over 75% year-on-year, the revenue of cloud service provider business increased by over 2.1 times year-on-year, and the revenue of cloud service provider GPU AI servers increased by over 5 times year-on-year, with a month on month increase of over 90%. In this regard, the company’s major clients are technology giants such as NVIDIA, Microsoft, Amazon, which are leading in AI and cloud services.

We have noticed that after the release of this strong performance, there have been continuous rumors in the market that Industrial Fulian’s future performance is not optimistic. Especially on November 12th, after Wistron, the core supplier of Nvidia AI GPU substrates and modules, released its financial report, some analysts suggested that Nvidia may switch to directly supplying Level-10 systems (i.e. complete server machines) to its partners in the Vera Rubin platform, which will begin mass shipments next year. As a midstream manufacturer and system integrator of Nvidia AI servers, especially Industrial Fulian, which has advantages in L10 and above systems, is believed to be significantly impacted.

Reflected in the stock price, the company’s stock price hit a high of 83.88 yuan on October 30th and then continued to decline. On November 24th, the closing price was only 55.94 yuan, a 33% drop from the high point. Within half a month, the market value plummeted from nearly 1.7 trillion yuan to 1.1 trillion yuan.

We have also noticed that as early as the performance briefing on November 13th, the issues of orders and performance in the fourth quarter and 2026 had already become the focus of attention regarding the recent small essays that can be considered as “financial losses” for Industrial Fulian. An analyst directly asked, ‘Recently, there have been rumors that the company has lowered its cabinet shipments and performance expectations. How should the current company understand this?’. The response from Industrial Fulian is that the company has not released any relevant statements regarding the rumors, and the progress and delivery pace of existing customer projects are normal. And it reminds that there is no need to overinterpret market rumors, stating that the company will “continue to expand diversified cooperation with customers, achieve stable delivery, improve profit quality, and achieve better results quarter by quarter”.

At the same time, Industrial Fulian also stated that the new product is progressing smoothly, with comprehensive upgrades in computing power density, power efficiency, and system reliability, especially in terms of power supply architecture, heat dissipation efficiency, and integration strength, bringing significant optimizations. In the future, related products will also become one of the core growth engines for the company’s high-end AI server delivery.

At the same time, there are also rumors circulating in some AI communities that Nvidia is entering the L10 or industry trend for standardized design and accelerated production, but it is currently unable to enter the L11 (rack level integration) and L12 (multi rack cluster integration) as some rumors suggest. Some institutions have also reported that there may be more opposition from the industry chain to Nvidia’s entry into L10, which means that at present, this is far from the truth.

In addition, industry insiders have pointed out that the main delivery products in the current performance of Industrial Fulian are GB200 and GB300, and it may not be until the second quarter of next year that Vera Rubin related products will be involved. Therefore, the statement of lowering the fourth quarter performance is clearly unfounded.

Industrial Fulian also clearly stated in investor communication that the company’s GB200 shipments have been very smooth, and GB300 has achieved mass production in the third quarter, with continuous improvement in yield and testing efficiency. The company expects a decrease in unit costs and an increase in yield to provide positive support for the gross profit margin in the fourth quarter.

 
© 版权声明
THE END
If you like it, please support it
点赞14 分享
comment Grab the sofa

Please log in and leave a comment

    No comment content available at the moment