Moore Threads, popularly known as the “Chinese Nvidia,” is about to debut on Shanghai’s STAR Market, bringing a mix of investor euphoria and still-red numbers. The company, specializing in general-purpose GPUs, has opened its IPO subscription period with a valuation that already places it among the major names in China’s new wave of AI chip manufacturers.
A record IPO on the STAR Market
According to the documents submitted to the market, Moore Threads will issue 70 million new shares at a price of 114.28 yuan per share, raising nearly 8 billion yuan and achieving a post-IPO valuation of approximately 53.7 billion yuan. This marks the largest debut of the year on the STAR Market, China’s Nasdaq-like index for high-growth tech companies.
Investor appetite has been intense from the early stages of the process. In the book-building phase, 267 institutions participated, with an over-subscription ratio of 1,572 to 1—one of the highest figures recorded in this market. The process also moved quickly: the company submitted its application at the end of June, passed regulatory review on September 26, and received final approval for the issuance in October. In total, just 122 days to complete the essential IPO procedures—a record time for the STAR Market.
China’s bet on a “homegrown” GPU
Moore Threads was founded with a clear goal: to offer a domestic alternative to Nvidia and AMD GPUs at a time when international export restrictions on advanced chips are pressuring China’s industry. Unlike other local accelerator manufacturers that have opted for ASIC designs tailored for AI, the company follows a general-purpose GPU model, offering more flexibility for graphics, scientific computing, and AI workloads.
Its own architecture, called MUSA, includes a key tool named Musify, designed to facilitate migration of code written for CUDA—the Nvidia ecosystem—toward Moore Threads’ platform. The goal is to lower the entry barriers for data centers and developers already working with Nvidia libraries who, for cost, availability, or regulatory reasons, want a local alternative.
In a recent investor conference ahead of the IPO, founder, chairman, and CEO Zhang Jianzhong summed up Moore Threads’ strengths: relative technological leadership within the Chinese GPU ecosystem, growing support for the MUSA architecture, strong integration with local software and hardware, and an R&D team with international experience.
It’s no coincidence that many talk about “Nvidia DNA.” Zhang spent 14 years at the American multinational, reaching the role of global vice president and head of Greater China. Several key executives at Moore Threads also held high-level positions at Nvidia before joining the project.
Recognizing the gaps with Nvidia and AMD
Despite market enthusiasm, Moore Threads openly acknowledges that a significant gap remains compared to global leaders. When asked about the differences with Nvidia during the investor conference, Zhang admitted the company needs to continue accumulating technology and improving product performance.
He explained that Nvidia has deep experience in GPUs and holds technical and cost advantages in performance, compatibility, and large-scale GPU clustering capabilities. Still, he defended that some Moore Threads products are already approaching or reaching advanced international levels in certain metrics, to the point of potentially replacing some “strategic” components.
For example, he cited that the FP32 computing power of the MTT S80 graphics card is comparable to a RTX 3060, and that the computational efficiency of its AI cluster, MTT S5000, with a thousand cards, surpasses foreign systems of the same generation and size used for training models.
A booming sector… and a red bottom line
The leadership’s optimism is backed by growth forecasts for China’s GPU market. According to estimates from Frost & Sullivan cited by the company, the market could reach 1,363.578 billion yuan by 2029, driven by the expansion of generative AI, high-performance computing, and data centers.
Based on this, Zhang is convinced that demand for computing infrastructure—and particularly GPUs—will experience explosive growth in the coming years, and Moore Threads is well-positioned to capture a significant share of that market.
However, behind the impressive revenue growth—compound annual rates exceeding 200%—financial statements reveal the less glamorous side typical of “hard tech” companies in heavy investment phase.
Moore Threads’ operating revenue jumped from just 460 million yuan in 2022 to 7.85 billion yuan in the first nine months of 2025. But losses have also soared: net attributable to shareholders was negative 1.894 billion yuan in 2022 and reached 7.24 billion yuan in the first three quarters of 2025. In total, nearly 6 billion yuan lost in just over three years and nine months.
CEO Zhang attributes these figures to the nature of the business: “The GPU industry is at a critical phase of technological breakthrough and market expansion. As a startup in China’s GPU chip sector, we remain in a key period of technological accumulation and market development, characterized by high R&D investments and the time needed to attract customers.”
Profitability horizon: not before 2027
One of the main concerns for future shareholders is the timeline for reaching profitability. Management has provided a clear outlook: based on market potential, product roadmap progress, and customer onboarding pace, Moore Threads expects to achieve consolidated profits no earlier than 2027.
In other words, the company asks investors for patience, as several years of additional losses are anticipated while the MUSA ecosystem is consolidated, sales scale up, and heavy investment in next-generation chip development is amortized.
The funds raised through the IPO will mainly be allocated to three major projects: fully autonomous next-generation AI training and inference chips, new GPUs developed in-house, and future AI System-on-Chips (SoCs), along with strengthening working capital to support business expansion.
A strategic bet amid the chip war
Moore Threads’ stock market debut occurs against the backdrop of increasing geopolitical and technological tensions. US and other countries’ restrictions on exporting advanced GPUs have accelerated China’s efforts to build a local supply chain that reduces dependence on foreign suppliers in critical areas like AI and high-performance computing.
In this context, Moore Threads embodies the bet on having a domestic general-purpose GPU supplier capable of competing—primarily within China—with Nvidia and AMD. The company’s success or failure will be closely watched by both Beijing authorities and major global industry players.
For investors, this IPO presents an opportunity to participate in one of China’s most ambitious growth stories in the chip ecosystem but also serves as a reminder of the risks associated with companies that, despite strong technological and political backing, are still far from demonstrating sustained profitability.
Frequently asked questions about Moore Threads’ IPO
Why is Moore Threads called the “Chinese Nvidia”?
Because its business model focuses on developing general-purpose GPUs, similar to Nvidia, rather than specialized ASIC chips. Additionally, many of its top executives, including its founder, come from Nvidia itself, reinforcing the comparison.
How does Moore Threads differ from other Chinese AI chip manufacturers?
While companies like Cambrian and Huawei’s HiSilicon division focus on ASIC designs tailored for AI, Moore Threads develops complete GPUs capable of handling graphics, AI workloads, and general computing. Its MUSA architecture and Musify tool aim to facilitate migration from Nvidia’s CUDA ecosystem.
What are its current financial results like?
The company is experiencing rapid revenue growth—from hundreds of millions to several billion yuan in just a few years. However, losses are also significant, with nearly 6 billion yuan in accumulated red ink over less than four years, mainly due to heavy investment in R&D and early-stage market expansion.
When does Moore Threads expect to turn a profit?
Management estimates that if the GPU market grows as expected and the company executes its technology and commercial roadmap, it could become profitable around 2027. Until then, it is expected to continue a phase of heavy investment and expansion, with associated risks but also growth potential aligned with China’s push for semiconductor independence.
via: news.futunn