Moore Threads, Hygon project triple-digit H1 growth as China's AI chip demand surges
Synopsis
Moore Threads projects up to 149.4% revenue growth in H1 2026 — a figure that underscores just how fast China's homegrown AI chip market is scaling in the absence of Nvidia hardware, with Hygon adding up to 70.2% growth on top of an already larger revenue base.
Key Takeaways
Moore Threads forecast H1 2026 revenue of between 1.65 billion yuan (US$243.5 million) and 1.75 billion yuan , representing growth of 135.1% to 149.4% year on year.
Hygon Information Technology projected H1 2026 revenue of between 8.5 billion yuan and 9.3 billion yuan , up 55.6% to 70.2% year on year.
Moore Threads credited its Kua'e AI computing clusters and its flagship MTT S5000 GPU — now in mass production — for the revenue surge.
Moore Threads listed on Shanghai 's Star Market (Sci-Tech Innovation Board) in December 2025 .
Both companies compete with Huawei Technologies ' Ascend chips and Cambricon Technologies for domestic AI infrastructure contracts.
Sustained US export controls on advanced chips have effectively created a captive domestic market, accelerating procurement from Chinese suppliers.
Chinese GPU maker Moore Threads Technology and CPU-and-accelerator designer Hygon Information Technology have each forecast massive revenue growth for the first half of 2026, signalling that domestic demand for AI computing infrastructure in China is accelerating sharply even as access to leading foreign chips remains constrained.
The Numbers
Moore Threads, headquartered in Beijing, stated in a stock exchange filing on Thursday, 17 July 2026, that it expected first-half revenue to rise between 135.1 per cent and 149.4 per cent year on year, reaching between 1.65 billion yuan (US$243.5 million) and 1.75 billion yuan. Hygon Information Technology projected a comparatively steadier — though still substantial — first-half revenue increase of 55.6 per cent to 70.2 per cent, with sales expected to land between 8.5 billion yuan and 9.3 billion yuan.
What Is Driving the Surge
Moore Threads attributed the outsized growth to robust demand for its full-function GPUs and the rapid commercial roll-out of its Kua'e AI computing clusters. The company also said its flagship chip, the MTT S5000, had achieved market-leading performance, is now in mass production, and — according to the company — delivers computing efficiency that matches top international alternatives. Hygon, meanwhile, designs both central processing units (CPUs) and deep computing units, a category of accelerator card built specifically for AI workloads, giving it exposure across multiple segments of the data-centre stack.
Why It Matters
The forecasts arrive against a backdrop of sustained US export controls that have restricted Chinese cloud providers and AI labs from procuring high-end chips from Nvidia, effectively creating a captive market for domestic alternatives. Both companies are widely regarded as part of Beijing's strategic push to build a self-sufficient semiconductor supply chain. Moore Threads went public in December on Shanghai's Star Market — officially the Sci-Tech Innovation Board — a listing venue designed specifically for high-tech firms.
The Competitive Backdrop
The two companies are not alone in racing to fill the gap left by restricted Nvidia and AMD hardware. Huawei Technologies' Ascend AI chips and Cambricon Technologies' inference accelerators are also competing for the same surge in domestic AI infrastructure spending. Industry analysts have noted that Chinese hyperscalers and state-backed enterprises are under pressure to demonstrate AI capability, creating urgent procurement cycles that benefit any credible domestic supplier regardless of absolute performance parity with global leaders.
What's Next
Full first-half results from both companies will provide a clearer picture of whether the forecast ranges are met — and whether margins hold as competition among domestic chip designers intensifies. Investors and policymakers in Washington and Beijing alike will be watching to gauge how quickly China's homegrown chip ecosystem can scale to meet enterprise-grade AI demand.
Point of View
Not just at the margins but at its core. What mainstream coverage often underplays is that Hygon's larger absolute revenue base, combined with its CPU-plus-accelerator portfolio, arguably makes it the more strategically significant company: it is addressing the full data-centre stack, not just GPU compute. The deeper pattern here is commoditisation under duress — Chinese buyers are not necessarily choosing domestic chips because they prefer them, but because they have little alternative, which means margin pressure for these suppliers could intensify the moment export-control regimes ease or domestic competition sharpens. Investors should watch whether either company can sustain growth rates once the initial pent-up demand wave is absorbed.
NationPress
17 Jul 2026
Frequently Asked Questions
What revenue growth did Moore Threads forecast for H1 2026?
Moore Threads projected H1 2026 revenue growth of between 135.1% and 149.4% year on year, with sales expected to reach between 1.65 billion yuan (US$243.5 million) and 1.75 billion yuan . The company attributed the surge to strong demand for its GPUs and the commercial roll-out of its Kua'e AI computing clusters .
What is Hygon Information Technology and what did it forecast?
Hygon Information Technology is a Chinese designer of CPUs and deep computing units — AI-specific accelerator cards. It projected H1 2026 revenue of between 8.5 billion yuan and 9.3 billion yuan , representing growth of 55.6% to 70.2% year on year.
Why are Chinese AI chip companies growing so fast?
Sustained US export controls have blocked Chinese firms from purchasing advanced chips from Nvidia and other foreign suppliers, creating strong captive demand for domestic alternatives. Chinese hyperscalers and state-backed enterprises are accelerating procurement from companies like Moore Threads , Hygon , Huawei Technologies , and Cambricon Technologies .
What is the Moore Threads MTT S5000?
The MTT S5000 is Moore Threads ' flagship GPU. According to the company, it has achieved market-leading performance and is now in mass production, with computing efficiency that the company claims matches top international alternatives.
Who are the main competitors to Moore Threads and Hygon in China?
The primary domestic rivals include Huawei Technologies , whose Ascend AI chips target the same enterprise and cloud market, and Cambricon Technologies , which focuses on AI inference accelerators. All are competing for the same surge in domestic AI infrastructure spending driven by restricted access to Nvidia hardware.
