Chinese artificial intelligence chipmakers, including Moore Threads and MetaX, reported profits in the third quarter of 2026 — a milestone for the domestic sector amid US export restrictions on advanced Nvidia processors. Growth comes from local demand for national alternatives, although high memory costs and supply-chain challenges limit margins.
What happened?
According to South China Morning Post reports and company results, Moore Threads recorded quarterly revenue of about 1.55 billion yuan (roughly 1,774% year-on-year growth) and its first net profit of around 49 million yuan. MetaX also posted a profit in the period, with revenue around 1.5 billion yuan. Other firms such as Iluvatar CoreX compete in the same space, seeking contracts with Chinese enterprise clients and cloud providers.
The push stems from the need to reduce dependence on high-performance foreign chips restricted by export controls. Companies like Cambricon had already shown stronger results, but the pure-play GPU startups mark the entry into profitability territory.
Why does this matter?
The development signals progress in China’s strategy for semiconductor self-sufficiency in AI. Analysts note, however, that the profits remain modest and vulnerable to high HBM and memory costs, plus capacity constraints at domestic foundries such as SMIC. Over the first nine months, some companies still show adjusted operating losses and heavy cash consumption.
What changes in practice?
Chinese buyers of AI infrastructure gain more local options, even if performance does not yet match Nvidia’s top-tier chips. For the global market, it reinforces the fragmentation of the AI chip supply chain. The challenge remains: scaling volume, improving efficiency and sustaining margins in an environment of expensive memory and restricted capacity.
Sources: South China Morning Post (November 6, 2026) and company financial reports.
By Grok
.jpg)
