MetaX prepares a launch in Hong Kong to finance its battle against Nvidia

MetaX aims to accelerate its transition from a rising Chinese GPU manufacturer to an artificial intelligence computing platform. The company, based in Shanghai and listed on the STAR Market since December 2025, has announced its intention to launch an H-share offering in Hong Kong less than six months after its initial public debut on the mainland technology exchange. The operation seeks to fund expansion, strengthen corporate governance, and advance its international strategy.

This move comes amid China’s fever for AI chips. U.S. restrictions on advanced semiconductors have increased pressure on Beijing to develop domestic alternatives to Nvidia, AMD, and other foreign suppliers. While MetaX is not yet a tech equivalent of Nvidia, it is part of the group of companies that China aims to develop as domestic bases for training, inference, graphics, and scientific computing.

A second financing route in less than six months

MetaX, with the Shanghai stock code 688802.SH, plans to issue H-shares representing no more than 5% of its enlarged capital after the offering, excluding over-allocation options. The company has not yet disclosed the pricing, final timetable, or the amount it expects to raise, but has indicated that the funds will be directed towards developing next-generation GPUs, supply chain investments, business expansion, and international growth.

This decision is significant because MetaX recently marked one of the most prominent debuts in the Chinese semiconductor market. The company went public on the STAR Market in December 2025, with an IPO aimed at financing research, commercialization, and ecosystem development. Its first day reflected strong investor interest in local AI chip manufacturers.

Hong Kong offers MetaX a more international investor base than the mainland market and greater visibility among funds tracking China’s semiconductor supply chain development. It would also help diversify funding sources at a stage when designing AI chips requires substantial investments and long development cycles.

The context: China’s pursuit of its own GPUs

MetaX’s strategy cannot be separated from the geopolitical landscape. China needs to accelerate self-sufficiency in AI chips because access to Nvidia’s most advanced GPUs remains subject to U.S. export controls. This pressure has sparked a wave of investment in companies like MetaX, Moore Threads, Biren, Enflame, and Kunlunxin, among others.

Demand exists. Data centers, cloud providers, generative models, industrial AI applications, local inference, and regulated sectors require computing capacity. But the technical challenge is enormous. Nvidia’s dominance is not just about the chip itself; it also stems from CUDA, libraries, developer ecosystems, support, framework integration, availability, performance, and software maturity.

MetaX aims to address these challenges with an integrated hardware and software strategy. The company develops complete GPU stacks and computing solutions supported by its MXMACA software. Its goal is to deliver chips and platforms capable of covering AI training and inference, graphics rendering, and scientific AI applications.

MXMACA, China’s effort to reduce dependence on foreign software

One of MetaX’s key elements is MXMACA, its proprietary software stack. In AI GPUs, software is not a mere complement — it’s part of the product. Without development tools, compilers, libraries, and framework compatibility, a chip might be limited even if it has attractive specifications on paper.

MetaX asserts that its co-design approach aims for efficiency and versatility. Since 2025, the company has been building its ecosystem under the principle of “open collaboration and independent control.” The political and industrial signals are clear: it seeks to build a computing base that isn’t wholly dependent on U.S. vendors.

The company has adopted a “1+6+X” strategy. The “1” represents digital computing infrastructure. The six main sectors are finance, healthcare, energy, education and research, transportation, and digital entertainment. The “X” encompasses emerging areas like embedded intelligence and low-altitude economy, where China sees new sources of technological demand.

It’s not just chips — it’s about winning clients

For MetaX, listing in Hong Kong can aid in funding product development, but the real test will be commercial. Selling AI GPUs is not just about manufacturing silicon; it requires convincing clients they can migrate workloads, train or run models, maintain stability, manage costs, and receive technical support. Often, shifting from Nvidia entails rewriting or adapting software.

The Chinese market provides an initial advantage. State enterprises, government agencies, universities, cloud providers, and regulated sectors might have incentives to adopt domestic solutions, even if their performance isn’t identical to that of international leaders. In a restricted environment, local availability can be as important as absolute performance.

International expansion will be more complex. Outside China, MetaX will need to compete with Nvidia, AMD, Intel, and others, as well as overcome doubts about performance, support, compatibility, sanctions, trust, and supply continuity. Hong Kong can provide greater visibility but doesn’t eliminate these challenges.

Valuation driven by technological sovereignty

MetaX’s debut in Shanghai showed how much Chinese investors are willing to pay for self-sufficiency in AI. The share price soared on its first day, aligning with enthusiasm for local GPU manufacturers. This euphoria stems from China’s need for national champions in chips, with markets seeking direct beneficiaries of this industrial policy.

However, enthusiasm also carries risks. Many of these companies are still in heavy investment phases, incurring losses, with high R&D expenses and products that must demonstrate maturity compared to global alternatives. Capital is essential, but it doesn’t guarantee execution.

MetaX is not only competing with Nvidia; it faces competition from other Chinese players like Biren, Moore Threads, Enflame, Huawei, and others vying for the same space. Access to capital will be crucial for funding talent, tape-outs, validation, software ecosystems, and customer relationships.

Hong Kong as a showcase for Chinese AI

This operation also signals something about Hong Kong. The market aims to position itself as a financing avenue for Chinese semiconductor and AI companies seeking capital beyond the mainland. For firms like MetaX, dual listing can combine domestic support with international exposure.

Interest in these companies has increased because AI chips are now a strategic priority. They are not just a technological category but critical infrastructure for generative models, defense, industry, cloud computing, research, and digital services. Investors recognize this, which explains their willingness to evaluate companies that a few years ago might have seemed too young for such high valuations.

Nevertheless, Hong Kong also demands greater transparency. The company will need to better articulate its roadmap, margins, supplier dependence, key clients, MXMACA’s compatibility, and real progress compared to competitors.

The real challenge: turning narrative into performance

MetaX’s case illustrates the current state of the Chinese semiconductor industry. There’s capital, political pressure, demand, and urgency. But there are also technical gaps, supply uncertainties, and fierce international competition. An IPO in Hong Kong might provide more resources but will not, by itself, solve the challenge of building a robust alternative to Nvidia.

For MetaX to succeed, it needs more than just funding. It must demonstrate performance in real AI workloads, software stability, product availability, developer support, and manufacturing capacity. It also must convince clients that its ecosystem is a long-term, reliable choice.

The planned Hong Kong operation confirms that China’s race for AI GPUs is becoming more financially driven. Companies are no longer only competing in labs and data centers; they are also competing for public capital, liquidity, investor confidence, and the ability to sustain years of investment.

MetaX aspires to be part of that national computation platform group. Its challenge is immense: it is not enough to be “China’s Nvidia challenger.” To survive, it must become a technically credible, commercially viable, and sufficiently open option to attract developers. The Hong Kong IPO can be a significant step, but the true race will be measured by performance, software, and adoption.

Frequently Asked Questions

What is MetaX?

MetaX is a Shanghai-based Chinese company specializing in GPU design, computing solutions, and software for AI, graphics, and scientific applications.

What are its plans in Hong Kong?

The company has announced its intention to issue up to 5% of its enlarged capital in Hong Kong through H-shares, prior to any over-allocation options.

Why compare MetaX to Nvidia?

Because it develops GPUs and computing platforms for AI training and inference, a market dominated globally by Nvidia. However, MetaX still needs to demonstrate scale, performance, and ecosystem maturity.

What is MXMACA?

MXMACA is the proprietary software stack developed by MetaX to complement its GPUs and facilitate an integrated hardware-software approach for AI workloads and general computing.