Baidu is rapidly emerging as one of China’s leading artificial intelligence chip contenders, positioning itself against Huawei as both companies attempt to fill the gap left by Nvidia’s restricted presence in the country. Best known for its dominance in China’s search engine market, Baidu has shifted its strategic focus in recent years toward autonomous driving and AI, largely through its majority owned chip subsidiary, Kunlunxin.
Analysts have recently upgraded their outlook on Baidu’s stock, citing its growing semiconductor efforts and projecting increased domestic demand for Kunlunxin products. This month, Baidu revealed a five year roadmap for its Kunlun AI chips, starting with the M100 chip in 2026 and the M300 in 2027. The company currently runs its ERNIE AI models using a combination of its own chips and Nvidia hardware.
Baidu monetizes its semiconductor technology by selling chips to third-party data center builders and offering cloud based computing capacity. The company aims to establish itself as a full-stack AI provider, encompassing chips, servers, data centers, AI models, and applications. Signs of momentum are already visible: earlier this year, Kunlunxin secured orders from suppliers to China Mobile, one of the country’s biggest carriers.
Deutsche Bank analysts recently noted that “Kunlunxin has emerged as a leading domestic AI chip developer, focusing on high performance chips for large language model training and inference, cloud computing, and telecom and enterprise workloads.” Nvidia’s GPUs are still considered the world’s most advanced for AI training and inference, but U.S. export restrictions prevent the sale of top end Nvidia chips to China. Beijing has also reportedly encouraged technology firms to avoid Nvidia’s weakened H20 chip, which is permitted for export but performs below U.S. restricted models.
With Huawei’s chip operations constrained, analysts believe Baidu is likely to fill a growing void. JPMorgan stated, “Domestic demand for AI compute in China remains intense, and hyperscalers are increasingly sourcing from local solution providers. We view Kunlun AI chip as one of the best positioned.” The bank forecasts Baidu’s chip revenue to jump sixfold, reaching 8 billion yuan ($1.1 billion) in 2026. Macquarie analysts estimate Kunlunxin’s standalone valuation could hit $28 billion.
Baidu is not the only tech giant building its own semiconductors. Alibaba is also developing next generation AI chips, while Tencent reports that chip shortages are limiting capital expenditure, not demand. Alibaba CEO Eddie Wu recently warned that supply constraints for key components will be a “large bottleneck” for two to three years.
China’s chip shortage is driven by both global semiconductor demand and U.S. restrictions on Nvidia’s advanced GPUs. Chinese companies have responded by using stockpiled chips and improving model efficiency to stretch available hardware. However, domestic manufacturing capacity remains limited. China’s leading foundry, SMIC, still lags behind Taiwan’s TSMC in both scale and technology, limiting China’s ability to produce enough high end chips to meet soaring AI demand.
Despite these challenges, AI demand in China remains extremely strong. “We are not even able to keep pace with the growth in customer demand,” Alibaba’s Wu said. This creates a major opening for Baidu. As Nick Patience of The Futurum Group explained, “Baidu’s chip push is both a necessity and an opportunity.” If Baidu can deliver competitive Kunlun chips on schedule, it could not only solve its own supply needs but also become a critical supplier for China’s wider AI ecosystem.
