China's AI Chips and Export Controls: Ascend, SMIC — 2026-10-26
Recent reports indicate a significant shift in China's AI chip market, with Huawei’s Ascend shipments nearly doubling to capture 28% of the market, while Nvidia’s share dropped to 34%. Cambricon reported strong Q3 financial results, with net profit rising 147%, as Beijing proposes new export controls on 23 AI technologies.
China's AI Chips and Export Controls: Ascend, SMIC — 2026-10-26
Top developments
Huawei Ascend Shipments Surpass Nvidia in Market Share
New data from IDC, cited by Reuters and local Chinese media, shows that Huawei’s Ascend AI chip shipments nearly doubled in the first half of 2026, capturing 28% of the Chinese market. This surge has pushed Nvidia’s market share down to 34%, a significant decline attributed to US export restrictions and Huawei’s domestic production ramp-up. The rise in Huawei’s share highlights the effectiveness of domestic substitution efforts despite ongoing supply chain constraints.
Cambricon Reports Record Q3 Profits
Cambricon, a key domestic AI chip designer, announced its Q3 2026 financial results on October 23, reporting a net profit of 1.404 billion yuan, a 147.74% year-over-year increase. Revenue for the quarter reached 3.531 billion yuan, up 104.47% from the previous year. For the first three quarters of 2026, Cambricon’s cumulative net profit hit 3.715 billion yuan, reflecting a 131.49% growth. These figures underscore the strong demand for domestic AI accelerators in China’s data centers.
Beijing Proposes Export Controls on 23 AI Technologies
China’s Commerce Ministry has proposed adding 23 artificial intelligence technologies to its export-control list, including AI model weights and chip designs. The public comment period for these new regulations closes on November 15, 2026. This move aims to tighten Beijing’s oversight of its burgeoning tech sector and potentially restrict the overseas transfer of advanced AI models and hardware.
Nvidia Locks in Production Capacity with Prepayments
To mitigate supply chain constraints for AI infrastructure, Nvidia is making significant prepayments to six key partners in the packaging, substrate, and co-packaged optics (CPO) sectors. CEO Jensen Huang noted that while market demand is immense, capacity supply remains a critical bottleneck. This strategy is designed to secure long-term component availability for future chip generations.
Local view
Local Chinese media outlets, including China Business Journal citing IDC data, are highlighting that non-GPU accelerators have surpassed GPU shipments for the first time in the Chinese data center AI market. Huawei’s share is reported at 28%, with domestic vendors collectively gaining ground against Nvidia’s 34%. Analysts note that while Huawei’s Ascend chips are gaining traction, the ecosystem still faces challenges in interconnectivity and advanced process nodes compared to Nvidia’s CUDA environment.
Context & numbers
- Huawei Ascend Market Share: 28% in H1 2026, up from previous years.
- Nvidia China Market Share: Dropped to 34% in H1 2026.
- Cambricon Q3 2026 Net Profit: 1.404 billion yuan (+147.74% YoY).
- Cambricon Q3 2026 Revenue: 3.531 billion yuan (+104.47% YoY).
- China Export Control Proposal: Covers 23 AI technologies, comment period ends Nov. 15, 2026.
On the radar
- Nov. 15, 2026: Deadline for public comments on China’s proposed AI export control list.
- Huawei Ascend 960: Reports indicate the next-generation Ascend 960 chip roadmap may be pulled forward to Q1 2027, though this news originated in September.
- SMIC Capacity Allocation: Ongoing discussions regarding SMIC’s advanced node capacity allocation among Huawei, Cambricon, and other domestic designers remain a key watchpoint for supply chain analysts.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
