What happened
The South China Morning Post reports that MetaX Integrated Circuits posted net profit of 612 million yuan for the first six months of 2026, reversing a loss of 186 million yuan a year earlier. Its revenue rose 44.7% to 1.32 billion yuan, according to a stock-exchange filing cited by the newspaper.
The South China Morning Post reports that MetaX Integrated Circuits earned 612 million yuan, or about US$90.9 million, during the first half of 2026. That compares with a loss of 186 million yuan in the same period a year earlier. The report describes the result as a swing to profitability by a Chinese graphics-processing-unit designer operating in the country’s increasingly important AI-chip market. The supplied source attributes the financial figures to a stock-exchange filing made on Sunday evening; the filing itself is not independently available in the material provided here.
SCMP also reports that MetaX’s first-half revenue increased 44.7% to 1.32 billion yuan. The company attributed the growth, as described by the newspaper, to widespread customer adoption and a sharp rise in GPU shipments. The report does not provide a breakdown of customers, shipment volumes, product lines, margins, or the portion of revenue directly related to AI workloads. Those missing details limit what can be concluded about the durability or composition of the reported growth.
The earnings report coincided with sharply different stock moves among Chinese AI-chip companies. SCMP reports that MetaX’s Shanghai-listed shares rose 1.4% to 684.5 yuan by Monday’s midday break. Biren Technology’s Hong Kong-listed shares rose 10.7% to HK$43.44, while Iluvatar Corex fell 6.7% to HK$370.2. The newspaper presents the moves as evidence of a widening profitability divide, but the supplied article does not establish that the companies’ share-price changes were caused solely by their financial results.
MetaX made its debut on Shanghai’s Star Market in December, according to SCMP. The newspaper identifies it as one of China’s most closely watched GPU developers and says it was founded in 2020 by former executives of U.S. chip company Advanced Micro Devices. The article places MetaX’s results within a broader national effort to build alternatives to Nvidia’s AI chips, but the supplied text contains no independent assessment of MetaX’s technical performance, market share, customer concentration, manufacturing arrangements, or competitiveness against foreign and domestic rivals.
Why it matters
The results offer a concrete indication that at least one Chinese GPU designer is converting demand for domestic AI computing hardware into reported profit. They also show that the country’s AI-chip sector is developing unevenly, with investor reactions differing significantly among companies.
The immediate significance is financial rather than merely promotional. A reported profit suggests that domestic AI-chip demand can, at least for one company and one reporting period, support more than development spending and market speculation. The South China Morning Post connects MetaX’s result to increased GPU shipments and customer adoption, providing a concrete business measure for China’s effort to develop alternatives to Nvidia’s hardware. The result does not by itself prove that the broader sector has become profitable or self-sustaining.
The figures also illustrate why the Chinese AI-chip market should not be treated as a single bloc. SCMP reports that three listed companies experienced materially different stock movements as investors assessed first-half results. That divergence matters for companies seeking capital, customers evaluating suppliers, and policymakers supporting domestic hardware capacity. It indicates that investors may be distinguishing among firms’ financial outcomes, even though the article does not provide enough information to identify the specific factors behind each stock move.
The story has broader infrastructure implications because GPUs are a foundational input for training and running AI systems. If domestic chip designers can increase shipments and generate profits, they may be better positioned to invest in production, software support, and future designs. However, the source does not report any new manufacturing capacity, major customer contract, benchmark result, export-control development, or deployment milestone. The practical significance therefore rests on the reported financial turnaround and its possible signal about demand, not on evidence of a technological breakthrough.
Important limitations remain. The supplied report does not independently confirm the exchange filing, provide Biren Technology’s or Iluvatar Corex’s comparable earnings figures, or explain whether MetaX’s profit includes unusual items. It also does not establish how much of the reported demand is recurring, how many customers are purchasing the GPUs, or whether the company can maintain margins as competition increases. Readers should treat the result as a reported first-half performance by one company, not as proof that China has closed the gap with Nvidia or solved its AI-compute supply constraints.
What to watch next
The key unanswered questions are whether MetaX can sustain its profit and shipment growth, whether Biren Technology and Iluvatar Corex report comparable financial performance, and how much of the sector’s demand depends on China’s state-backed push for technological self-sufficiency and AI infrastructure.
The next useful evidence would be fuller financial disclosures from MetaX and comparable first-half reports from Biren Technology and Iluvatar Corex. Those documents could show whether the reported profitability divide reflects differences in revenue growth, gross margins, operating costs, customer concentration, shipment scale, or accounting treatment. The supplied SCMP report names the divergent stock moves but does not provide those underlying comparisons.
Investors and industry observers should also watch whether MetaX’s revenue growth continues after the reported half-year period. SCMP attributes the increase to customer adoption and higher GPU shipments, but the article gives no information about repeat orders, backlog, cancellations, pricing, or the balance between government-linked and commercial demand. Without those indicators, it is not possible to determine whether the profit represents a durable operating trend or a short-term phase of the domestic infrastructure buildout.
A further question is how the companies perform in actual AI deployments. The source does not include independent benchmarks, customer satisfaction data, software-compatibility information, energy-efficiency comparisons, or evidence from large-scale production use. Such information would help distinguish financial momentum from practical competitiveness. The broader policy environment also merits attention because SCMP links the companies’ prospects to Beijing’s push for technological self-sufficiency and national AI infrastructure, but the article does not identify specific subsidies, procurement decisions, or policy changes driving the results.
For now, the clearest development is narrow but consequential: SCMP reports that MetaX moved from a first-half loss to a substantial profit while its revenue and GPU shipments increased, amid sharply mixed trading in peer companies. What remains unknown is whether that performance can be repeated, whether peers show similar economics, and whether reported profitability will translate into reliable, competitive AI hardware for customers.