A wave of initial public offerings (IPOs) is sweeping China’s graphics processing unit (GPU) sector, marking a critical phase in the nation’s push for self-reliance in advanced computing power. On January 22, Shanghai-based Enflame Technology’s application to list on Shanghai’s Nasdaq-style STAR Market was accepted by the exchange. This move paves the way for “The Four Dragons” – Enflame, Moore Threads, MetaX and Biren Technology – to complete their entries in the capital markets. This follows the successful listing of another Shanghai GPU firm, Iluvatar CoreX, on the main board of the Hong Kong stock exchange on January 8. It became the second home-grown GPU company to go public in Hong Kong after Biren. Earlier, industry leaders Moore Threads and MetaX had already debuted on the STAR Market to significant investor interest.
The listing momentum which started from late 2025 extends beyond dedicated chip designers. Baidu has announced plans to spin off its non-wholly owned subsidiary Kunlunxin for an independent listing. Informed sources told China Daily that Alibaba Group is also considering a similar spin-off plan for its chip design arm T-Head. This IPO rush unfolds against a backdrop of stringent U.S. semiconductor export controls and an explosive growth in domestic artificial intelligence applications, thus driving demand for alternative computing solutions.
At a recent news conference, Li Chao, Spokeswoman for the National Development and Reform Commission (NDRC), highlighted China’s deepening “AI+” initiative, which “provides extensive application scenarios for AI computing chips”. Li pointed to “rapidly growing demand and vibrant innovation across all types of computing chips”, adding that “domestic chip products are accelerating their adaptation across different scenarios with very good results”. The comments acknowledge the sector’s vitality while setting higher expectations for its future development. Computing power forms the bedrock of the digital economy, a strategic resource underpinning data flow, algorithm training and intelligent application deployment. If AI is a lush crop, computing infrastructure is the “fertile black soil” it depends on. As the global computing industry shifts from scale expansion to qualitative leaps, building a secure ecosystem, from underlying chips to applications, has become imperative for China to secure its developmental sovereignty, experts said.
Dong Peng, Senior Economist and Member of the Assets Committee of the China Enterprise Confederation, described this listing wave as “a collective breakthrough for domestic computing power during a strategic window of opportunity”. “Beneath the surface heat in capital markets lie two fundamental drivers,” Dong said. “First, the national demand for autonomy and controllability in AI and high-end manufacturing has created an unprecedented dual engine of policy and market sup- port. Second, Washington’s export restrictions on U.S. leaders like Nvidia have opened a precious market window for domestic firms.” He emphasized that listing is not merely about fundraising, but a crucial step to establish industry standards and ecosystem influence. “The race to go public is about seizing the advantage in technology iteration, customer lock-in, and brand recognition while the competition landscape is still taking shape. This is a battle for survival and, more importantly, for future power.”
The Hong Kong stock exchange and the STAR Market, as preferred boards for “hard tech” firms, have provided a streamlined pathway for these companies. Driven by continuous investment in AI infrastructure and the escalating strategic priority of technological self-reliance, China’s domestic AI chip sector is poised for substantial market expansion, experts said. Zhou Di, Senior Engineer and expert with the National Science and Technology Expert Database of the Ministry of Science and Technology (MST), said China’s AI computing market is likely to evolve toward a “dual-track, domestically led” new market landscape. “Nvidia is likely to retain a share in the high-end AI training market, represented by the demand of hyper-scale internet companies and cutting-edge research institutes, by virtue of its technical advantages. Domestic GPUs, leveraging cost-performance ratios, customized services, policy support and security advantages, are positioned to dominate vertical sectors like government affairs, finance, industry and healthcare, as well as broader cost-performance-sensitive markets,” Zhou said.
The wave of listings also comes as China’s push for self-sufficiency in AI processors is gaining remarkable momentum, with at least nine domestic chip companies having surpassed a significant threshold of 10,000 units in terms of order shipments. This 10,000-chip club signals that domestic AI chips are gaining tangible market traction based on their performance, stability and cost. It heralds a shift from mere scale competition to a more comprehensive battle encompassing software ecosystems, commercial services and sustained reliability, experts said, as reported by the China Daily.
