Moore, Mu Xi, BIREN TECH, and Tian Shu! The "Four Little Dragons of Domestic GPUs" gather for IPO

In today's fiercely competitive global computing power race, China's GPU industry is experiencing a historic collective breakthrough moment.

Following the listing of Moore Threads on the Sci-Tech Innovation Board with a market value exceeding 300 billion, Muxi Co., Ltd. has followed suit with a strong listing, BIREN TECH has just passed the Hong Kong Stock Exchange hearing, and TianShu ZhiXin has submitted its prospectus and is poised for launch— the "Four Little Dragons" of domestic GPUs have gathered in the capital market.

Although the four companies are in the same race, their differentiated paths are already very clear: from pursuing a platform strategy with a full-function ecosystem, to focusing on engineering capabilities for single-quarter profitability, to binding with national-level computing power hubs in a cluster route, each company is seizing market share in the chip market through different approaches.

In this intense competition, the differentiated competitive paths of each company have become clear. TianShu ZhiXin, leveraging its first-mover advantage of "earliest mass production, most stable supply," has achieved diversification in its customer structure and steady revenue growth through its two mature product lines, TianGai (training) and ZhiKai (inference).

Meanwhile, Moore Threads, which has just landed on the Sci-Tech Innovation Board, is waving the banner of "ecosystem supremacy," using the MUSA architecture's full-function GPU strategy to bridge the B-end and C-end markets. Its grand ecological layout, comparable to NVIDIA, and market value exceeding 300 billion demonstrate the high expectations of capital for its platform-level value.

At the same time, the other two giants are showing strong momentum in computing power infrastructure and commercialization efficiency. Muxi Co., Ltd. has completed a stunning listing with a market value of 200 billion at an astonishing speed and is set to achieve single-quarter profitability in the second quarter of 2025, becoming the fastest runner among the Four Little Dragons to break the industry's loss norm and realize commercialization.

In contrast, BIREN TECH, which is about to become the "first domestic GPU stock on the Hong Kong Stock Exchange," showcases its strength backed by a "national team" capital lineup in large-scale cluster deployment and cooperation with operators.

TianShu ZhiXin: First-Mover Advantage + Full Product Line, Rapid Decline in Customer Concentration

As one of the earliest companies in China to initiate general-purpose GPU design, TianShu ZhiXin has established a robust position in the supply chain with its first-mover advantage of "earliest mass production, most stable supply." According to its prospectus, the company is the first in China to achieve dual mass production of "training + inference" general-purpose GPUs and is the first Chinese design company to reach this milestone using 7nm technology.

The completeness of the product line is TianShu ZhiXin's core moat. Currently, the company has two mature product lines, TianGai (training) and ZhiKai (inference), covering AI model training and inference scenarios, thereby constructing a complete general-purpose GPU product portfolio, with cumulative shipments exceeding 53,000 units.

This dual-driven product strategy enables it to cover the full range of needs from complex model development to efficient production deployment. Financial data shows that over half of its revenue in the past three and a half years has come from high-priced training series products, demonstrating its competitiveness in the core computing power market The prospectus shows that the company's revenue is experiencing rapid growth, with revenue reaching 324 million yuan in the first half of 2025, surpassing the total for the entire year of 2023. At the same time, the company's customer structure has significantly optimized, with a substantial reduction in reliance on single customers. In 2022, the revenue share of the top five customers was as high as 94.2%, but by the first half of 2025, this ratio had significantly decreased to 38.6%, and the total number of customers grew from 22 to 106.

The rapid decline in customer concentration indicates that TianShuZhiXin has freed itself from reliance on a few early large customers, and the versatility of its products and market recognition are rapidly penetrating across the industry. In addition, more than one-third of the R&D team has over ten years of experience, and the executive team includes senior experts from international giants like Nvidia and AMD, further ensuring the stability of technological iteration.

Moore Threads: Ecosystem is King, Platform Value is Promising

Moore Threads, which has just landed on the Sci-Tech Innovation Board, has entered the forefront of the sector with a market value exceeding 300 billion yuan, with the market's high valuation largely paying for its vast ecological imagination.

Unlike vendors that focus solely on a single computing scenario, Moore Threads emphasizes a full-function GPU strategy. The company has launched four generations of GPUs based on its self-developed MUSA architecture, achieving breakthroughs in single-chip support for AI computing, graphics rendering, and scientific computing simultaneously. Its MTT S80 graphics card is close to Nvidia's RTX3060 in single-precision floating-point computing power, with some performance indicators reaching internationally advanced levels.

This full-function layout allows it to cover diverse needs from AI intelligent computing to personal computers. The prospectus shows that in the first half of 2025, the company has sold five AI clusters, demonstrating its product competitiveness and substitution potential in "bottleneck" areas such as intelligent computing centers.

Although it has not yet turned a profit, Moore Threads shows strong growth momentum. The company expects revenue to reach up to 1.498 billion yuan in 2025, with a year-on-year growth rate exceeding 240%, and the overall gross margin has improved to around 70%. According to its plan, the company expects to achieve profitability as early as 2027.

Investors' high valuation of Moore Threads essentially reflects their confidence in its long-term ecological imagination. As one of the few domestic GPU companies that simultaneously targets both B-end and C-end markets and supports full computing precision, once an ecological closed loop is formed due to widespread application adaptation, its platform will have a strong amplification effect. This "ecosystem is king" path is a core embodiment of the growth logic that parallels Nvidia.

BIREN TECH: Backed by the National Team, First Choice for Computing Power Hub

BIREN TECH, which has passed the Hong Kong Stock Exchange hearing, has taken a distinctive path deeply tied to the "national-level computing power upgrade dividend." On December 17th, according to the Hong Kong Stock Exchange, BIREN TECH is set to become the first "domestic GPU stock" in the Hong Kong market after passing the hearing. The prospectus shows that the company's revenue will rise to 337 million yuan in 2024, with a compound annual growth rate of up to 2500% over two years.

The company is backed by a luxurious "national team" of capital, including Shanghai Guotou Pioneer Fund, Guangzhou Industrial Investment, and Ping An Group, which gives it a natural advantage in undertaking national-level computing power hub and supercomputing center projects.

In terms of technology, BIREN TECH focuses on kilowatt-level clusters, Chiplet technology, and optical interconnect technology, concentrating on solving system-level challenges in large-scale computing power deployment. Data shows that in 2024, the company established deep partnerships with the three major telecommunications operators and successfully delivered a 1024-GPU intelligent computing cluster project worth 180 million yuan. This ability to deliver large orders demonstrates the maturity of its products in telecom-grade high-reliability scenarios.

Although the revenue scale is still in the climbing phase, BIREN TECH's growth is explosive, with a compound annual growth rate of 2500%. As the application of the Bili series products deepens in AI large model training and inference, especially with the rapid adaptation to cutting-edge models like DeepSeek, BIREN TECH is gradually establishing its position as a core supplier of computing power infrastructure.

Muxi Co., Ltd.: Fastest commercialization, first to achieve quarterly profitability

Compared to the other three companies, Muxi Co., Ltd. has demonstrated strong engineering capabilities and commercialization speed. This company, with deep AMD technical genes, experienced a commercial explosion just four years after its establishment, breaking the industry's continuous loss pattern.

Information on December 17th shows that Muxi Co., Ltd. opened at 700.00 yuan, soaring 568.83% from the issue price, with a current total market value of 281.4 billion yuan, becoming the second domestic GPU company to go public after Moore Threads. As a national-level specialized and innovative "little giant," it took less than six months from IPO acceptance to listing.

More importantly, Muxi Co., Ltd. has already welcomed its financial turning point—by the second quarter of 2025, the company achieved a net profit attributable to the parent company of 46.6189 million yuan, becoming the first among the four little dragons to achieve quarterly profitability. From January to September 2025, the loss narrowed significantly, achieving a net profit of -346 million yuan.

Muxi Co., Ltd.'s approach is pragmatic and efficient. The company does not blindly pursue concepts but relies on the strong product implementation capabilities of the AMD team to quickly launch three major product lines covering AI computing, general computing, and graphics rendering. It not only achieved several times year-on-year revenue growth in the first half of 2025 but also expects annual revenue to exceed 1.5 billion yuan The company's efficient execution capability stems from its technical team background, with core executives Chen Weiliang, Peng Li, Yang Jian, and others having long tenures at AMD Shanghai, possessing strong product engineering implementation capabilities. This technical accumulation has driven the rapid iteration of three major product lines: the domestically-sourced Xiyun C600 chip is expected to enter risk mass production by the end of the year, while the flagship product Xiyun C700 aims to compete with NVIDIA's H100. Currently, the company's cumulative GPU sales have exceeded 25,000 units and have been deployed in several national and operator-level intelligent computing platforms