Loading story…
China has initiated a policy shift to promote domestic AI hardware suppliers, officially assembling a government-approved list that includes only Cambricon and Huawei, while excluding foreign companies such as Nvidia and AMD. This move is designed to encourage public sector organizations to prioritize locally developed AI processors, with the updated Information Technology Innovation List (Xinchuang) now encompassing domestic AI accelerators alongside other homegrown technologies. The policy aims to accelerate the displacement of U.S.-designed AI hardware within the state sector, directing billions of dollars in government procurement toward Chinese suppliers and reducing reliance on foreign technology.
Cambricon Technologies is responding to this policy environment by planning a significant production ramp-up, targeting the delivery of 500,000 AI accelerators in 2026—more than triple its 2025 output. The company’s Siyuan 590 and 690 processors are central to this effort, with major customers like ByteDance and anticipated expansion to Alibaba. However, Cambricon faces challenges due to manufacturing constraints, as it relies on SMIC’s N+2 (7nm) process, which uses DUV lithography and has lower yields compared to leading global foundries. The government’s push for domestic hardware is expected to further shape the competitive landscape and drive rapid development of China’s AI ecosystem, despite technical and supply chain hurdles.
See the reporting duties and controls this puts on the clock.
3 events from the most recent confirmed update back to the earliest known activity.
Cambricon disclosed plans to deliver about 500,000 AI accelerators in 2026, including roughly 300,000 units of its Siyuan 590 and 690 processors. The production goal reflects rising demand from Chinese customers such as ByteDance and Alibaba and the government's push for domestic AI hardware.
Chinese authorities began offering substantial energy subsidies to cloud providers that use domestically produced AI hardware, aiming to accelerate replacement of U.S.-designed accelerators. The policy sought to boost uptake despite domestic chips lagging Nvidia in performance, software ecosystem, and manufacturing scale.
China expanded its Information Technology Innovation List (Xinchuang) to include AI hardware, with Cambricon and Huawei appearing as approved suppliers for public-sector procurement. The move is intended to steer government organizations toward domestically developed AI processors and away from foreign products such as Nvidia and AMD.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
See what this changes for your reporting obligations and which controls it puts on the clock.
2 references tracked. Mallory keeps watching after this page renders.
tomshardware.com
Open sourcetomshardware.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
China has implemented a comprehensive ban on the use of foreign AI chips, including those from Nvidia, AMD, and Intel, in any data center projects funded by the government. The directive, which reportedly applies retroactively to projects still in early stages, requires the removal or replacement of foreign accelerators even if they have already been installed. Chinese regulators have instructed that only domestically produced silicon, such as chips from Huawei, Cambricon, and Enflame, may be used in these environments. This move marks a significant escalation from previous policies that merely discouraged foreign chip purchases, and it effectively eliminates the prospects for foreign chipmakers in China's state-backed AI infrastructure. To support this transition and further reduce reliance on foreign technology, local Chinese governments are offering substantial power incentives and subsidies to domestic tech companies like ByteDance, Alibaba, and Tencent. These measures are designed to offset the higher energy costs and lower efficiency of Chinese-made AI chips compared to leading foreign alternatives, particularly Nvidia GPUs. Despite the technological gap, China is accelerating the development and deployment of homegrown AI hardware, with the government leveraging both regulatory and economic tools to promote domestic innovation and self-sufficiency in the semiconductor sector.
Jul 12, 2026Huawei and Cambricon are increasing production of AI accelerators at Chinese fabs, aiming for over a million units in 2026. However, production is constrained by limited advanced semiconductor fab capacity and a looming shortage of high-bandwidth memory (HBM). These bottlenecks threaten China's ambitions for AI hardware self-sufficiency.
Jun 29, 2026
China reportedly began approving imports of high-end *Nvidia* AI GPUs (including the **H200**) after weeks of uncertainty, with initial licenses expected to prioritize major Chinese internet companies building large AI data centers. Reporting cited by Reuters and the *South China Morning Post* indicates Beijing is attaching restrictive, not-yet-finalized conditions to these licenses—potentially including requirements to bundle purchases with domestic chips—reflecting an effort to balance near-term demand for leading accelerators with longer-term goals of strengthening China’s indigenous semiconductor ecosystem. In parallel, Chinese leader **Xi Jinping** publicly framed AI as an “epoch-making” technological transformation and called for faster progress in domestic development via a “whole-of-nation” approach, positioning AI as central to China’s upcoming 15th Five-Year Plan (through 2030). While not a discrete cybersecurity incident, the combined reporting is relevant to security leaders because it signals continued geopolitical pressure on AI supply chains and potential compliance, procurement, and third-party risk impacts for organizations operating in or selling into China’s AI infrastructure market.
Jun 29, 2026