In recent days, just 7 days after its debut on the A-share market, “Maoer Thread”, engaged in GPU design and research, announced that it will allocate up to 7.5 billion yuan out of the raised 75.8 billion yuan for wealth management. The news sparked continuous condemnation, turning the investment sentiment from enthusiasm to questioning and concern.
Many are shocked that “Maoer Thread” has idle funds for wealth management. Some analysts believe that under the backdrop of China’s AI national strategy, where capital and government trust are lacking, and businesses have never been trustworthy, only those who are being taken advantage of would believe such actions. The sentiment in the stock market is seen as a conspiracy.
On the evening of December 12th, “Maoer Thread”, known as the “first A-share of domestic GPU”, announced that it would use some idle raised funds for cash management, planning to use up to 7.5 billion yuan for a period of up to 12 months. The investment types include but are not limited to agreement deposits, notice deposits, fixed-term deposits, structured deposits, large-sum certificates of deposit, and other high-security, highly liquid, capital-protected products.
“Maoer Thread” landed on the A-share Sci-Tech Innovation Board on December 5th, with the stock price skyrocketing to 941 yuan within just five trading days, with a cumulative increase of over seven times, and the market value exceeding 440 billion yuan. So far, it has raised 75.76 billion yuan for chip development and to supplement working capital. However, allocating up to 75 billion yuan for wealth management has sparked controversy.
The public opinion reacted strongly to the news, with netizens mocking, “The end of chips leads to wealth management,” and even joking, “Huang Renxun would laugh out loud if he saw this.”
Among the 335,000 fans on YouTube channel “Xiao Cui’s Politics and Economics,” YouTuber Xiao Cui expressed two doubts in the “Xiao Cui Politics and Economics” program.
Firstly, the GPU industry is one of the most capital-intensive and long-term investment races in the world, a systematic mega project. It requires long-term investment in architecture, continuous evolution of instruction sets, and year-round spending on compiler and driver software ecosystems. Each generation of the product undergoes high-risk and high-cost tape-out and verification. Is “Maoer Thread” really in a critical stage of technological competition?
Raising a total of 75.8 billion in funds, yet having 75 billion idle for wealth management raises suspicions. It gives reasons to doubt whether “Maoer Thread” is really in a crucial stage of project research and development, or if it’s just an empty shell company!
Secondly, Xiao Cui stated that with “Maoer Thread” showing such relaxed development investment, its stock price should not have been pushed so high. The logic is simple, with such slow development investment from the company, why should the market price it based on the standard of the successful domestic Nvidia in the future?
However, many others believe that using shareholders’ money for wealth management violates the commitments made during the IPO and disappoints investors’ expectations.
New media “Top Financials” wrote that while boasting domestic substitution during fundraising, “Maoer Thread” seemed more interested in wealth management after going public, failing to withstand the test of its original intentions when it comes to funding. The battlefield for chips is about time, while wealth management is about interest gains. Going the wrong direction, even more effort would be in vain.
Blogger “Ride Cow Yan” said that the reason the market gave “Maoer Thread” such a high valuation, even if making losses, was to urge it to focus on research and development rather than deposit accounts! Just this one issue has undoubtedly tarnished the reputation of “Maoer Thread” in my eyes!
Blogger “Giant Stone Viewpoint” shared in a video, saying, “Shareholders provide you funding hoping you’ll be at the forefront, overcoming challenges, but before the battle has even begun, you’ve already exchanged your ammunition for two fixed deposits. This is not only absurd but also the biggest betrayal of the underlying capital market. If you need to invest in wealth management, do you really need to do it? Can’t the shareholders do it themselves?”
He further stated, “The fundraising capability of Chinese listed companies far exceeds their research and development capabilities. Behind the idle 75 billion in funds, it’s not simply the company’s unpredictability, but a genuine strategic confusion. Leading players are tossing huge amounts of raised funds into banks, earning interest from deposit accounts.
“This is not just ‘Maoer’s’ problem, many unicorn companies have this common flaw. They tell stories before going public, exaggerating their valuations, but after going public, they have nothing to do and turn funding into wealth management. Hardcore tech companies end up becoming inefficient financial intermediaries.”
Some netizens expressed, “Since ‘Maoer’ doesn’t need so much capital, why raise such a large funding all at once? Why not raise multiple times to allow the market to price more accurately and reduce the risk? This is not just a business issue but also a regulatory issue.”
According to “Maoer Thread’s” prospectus, the three chip research and development projects include AI training and reasoning chip development project, graphics chip development project, and AISoC chip research and development project. The company plans to invest approximately 25.1 billion yuan, 25 billion yuan, and 19.8 billion yuan respectively in these projects, with an additional 10 billion yuan to be used for working capital.
“Maoer Thread” is a general GPU company. “General-Purpose GPU” (GPGPU) refers to using the powerful parallel computing capabilities of Graphics Processing Units (GPUs) for general scientific calculations, data processing, AI, and other tasks originally processed by Central Processing Units (CPUs).
In September this year, the Communist Party officially announced the full launch of the “AI+”, transitioning AI from an auxiliary tool to a strong driver of economic growth. This strategic initiative has set a roadmap for the next ten years, focusing on promoting AI empowerment in industrial manufacturing, consumption, and other sectors to achieve the transformation from “enhancing efficiency” to “creating value”.
Under the promotion of the Communist Party’s AI national strategy, GPU chip research companies such as Muxi Technology, Biren Technology, and Suiyuan Technology are rushing to go public.
On the fifth day after landing on the A-share market, “Maoer Thread” briefly surpassed 900 yuan, becoming the third-highest A-share company after Guizhou Maotai and Horizon. The closing total market value exceeded 440 billion yuan. Despite this, the financial reports indicate that “Maoer Thread” is still a startup company. It generated revenue of only 785 million yuan in the first three quarters of 2025, but it has already surpassed the total revenue of the three years from 2022 to 2024.
From 2022 to 2024 and the first three quarters of 2025, the company suffered net losses of 1.894 billion yuan, 1.703 billion yuan, 1.618 billion yuan, and 724 million yuan respectively, with a combined net loss of 5.939 billion yuan over the three years and nine months.
Xiao Cui analyzed that the A-share market has become distorted, applying the valuation logic of top global AI giant Nvidia to price a newly established, heavily cash-burning domestic GPU company. However, this practice is a standard process in the A-share market, a common routine, except this time it is capitalizing on the national AI narrative.
“Maoer Thread” is simply taking advantage of the policy trend to raise funds first and decide how to use them later. If there are no projects available, going public is the ultimate goal of such companies. As for outpacing the competition on the curve and contributing to domestic GPU development, it’s better to just listen to it and not take it seriously.
Xiao Cui believes that any national strategy in the A-share market will undergo financialization, creating a mechanism for distribution. It’s essentially a conspiracy.
She said, “In a healthy market, market value rises with fundamentals, but in the A-share market, it’s the opposite, with stock prices rising first, and companies then playing catch-up with the fundamentals. For a company like ‘Maoer Thread’ with decades of growth potential, it’s absurd that it could be prematurely justified within 5 days. This makes the company lose the motivation to catch up with the fundamentals. Valuation is already in place, so why bother trying to innovate?”
“Only the ‘leeks’ in the stock market would believe in a national strategy or a curve overtaking narrative. Capital doesn’t believe it, the government doesn’t believe it, and the companies themselves have never believed it. They have just cooperated to create a scam.”
It’s worth noting that domestic GPU chip companies are facing two major hurdles on the technology front.
On December 15th, the International Cooperation Department of the China Securities Regulatory Commission issued a filing notice regarding Shanghai Biren Technology Co., Ltd.’s overseas issuance and listing and the “full circulation” of domestically unlisted shares, with Biren Technology planning to issue no more than 372,458,000 (approximately 3.72 billion) shares of overseas-listed common stock and to list on the Hong Kong Stock Exchange.
In addition, Muxi Technology will debut on the Sci-Tech Innovation Board for listing on December 17th, while Suiyuan Technology is currently in the IPO assistance stage.
The new media “Top Financials” also mentioned that one of the bigger challenges that “Maoer Thread” is facing is the dual bottleneck of software ecosystem construction and supply chain production capacity.
Firstly, Nvidia’s monopolistic position stems not only from its leading chip hardware but also because it has built a CUDA ecosystem for over two decades. Millions of developers worldwide write code based on the CUDA framework, creating a closed loop of “hardware-software-developers”, making it challenging for new entrants to outperform in the short term.
CUDA (Compute Unified Devices Architecture) is a hardware-software integrated technology introduced by Nvidia, providing a computational platform for applications.
For “Maoer Thread” to make a breakthrough, it must build its own MUSA architecture ecosystem – convincing software vendors to adapt, offering courses in universities to train developers, accumulating code in open-source communities, among other tasks.
On the other hand, the global high-end chip manufacturing, packaging, and testing capacities are highly concentrated, and domestic companies face the embarrassment of “having the money but unable to queue up.” For example, in CoWoS packaging, a critical stage for high-end AI chips, its capacity is predominantly occupied by international giants like Nvidia and AMD. Domestic manufacturers often have to wait for over six months to secure capacity.
CoWoS (Chip-on-Wafer-on-Substrate) is an advanced packaging technology by TSMC, crucial for AI chips (like NVIDIA GPUs).
By addressing these challenges and taking strategic steps, “Maoer Thread” and other domestic GPU chip companies aim to compete on a global scale and contribute to China’s technological advancement in the field of AI.