Company positions Kunlunxin unit to meet domestic demand for AI compute
Baidu is becoming one of China’s major artificial intelligence chip suppliers as domestic companies look for alternatives to Nvidia, whose most advanced GPUs remain restricted from the Chinese market. Known primarily for its search business, Baidu has shifted its focus toward driverless cars and AI, supported by Kunlunxin, a majority-owned subsidiary that designs AI chips.
Several analysts have upgraded their outlook on Baidu in recent weeks, citing momentum in its semiconductor business and anticipating increased domestic orders. Earlier this month, Baidu released a five-year roadmap for its Kunlun AI chips, starting with the M100 in 2026 and the M300 in 2027. The company already uses a combination of in-house chips and Nvidia processors in its data centers to run its ERNIE AI models.
Chip sales and cloud rentals form core of Baidu’s AI infrastructure push
Baidu sells its chips to companies building data centers and also provides computing capacity through its cloud business. It markets itself as a “full stack” AI provider, integrating chips, servers, data centers, AI models, and applications. Kunlunxin has recently gained traction, including orders from suppliers to China Mobile.
Analysts at Deutsche Bank noted that Kunlunxin has become a leading domestic developer of high-performance AI chips for large language model training and inference, cloud workloads, and telecom and enterprise systems.
Nvidia’s absence and Huawei’s constraints create opportunity
Nvidia’s top-end GPUs are blocked from sale to China under U.S. export controls, and Beijing has reportedly encouraged companies not to purchase Nvidia’s H20 chip, which is permitted for export but less powerful. Analysts say Baidu’s chip strategy is benefiting from this gap, particularly as Huawei’s large domestic clusters are affected by supply limitations.
JPMorgan said domestic demand for AI compute remains strong and that hyperscalers are increasingly sourcing from local providers. The bank forecast Baidu’s chip revenue could reach 8 billion yuan ($1.1 billion) in 2026, a sixfold increase. Macquarie estimated Kunlunxin could be valued at about $28 billion.
Industry-wide chip shortages shape China’s AI expansion
Alibaba and Tencent have also reported supply constraints. Alibaba CEO Eddie Wu said components and chips will be a bottleneck for the next two to three years, while Tencent President Martin Lau said the company’s revised 2025 capital expenditure plan reflects chip availability rather than weaker AI demand.
Companies have attempted to address shortages by using stockpiled Nvidia chips and improving model efficiency. However, China faces manufacturing limitations because its largest chipmaker, SMIC, cannot match the scale or technology of Taiwan Semiconductor Manufacturing Co.
Despite these constraints, demand for AI compute continues to rise. Wu said Alibaba cannot deploy new servers fast enough to keep up with customer requirements.
Analysts say Baidu could become a strategic supplier if its roadmap stays on schedule
Nick Patience, practice lead for AI at The Futurum Group, said Baidu’s expansion is both a necessity and an opportunity. He noted that U.S. GPU restrictions have created a large domestic market for compliant AI hardware aligned with Beijing’s self-reliance goals. If Baidu delivers competitive Kunlun chip generations on time, analysts say it could address its own supply challenges while becoming a central supplier to China’s AI ecosystem.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.