In recent months, executives in China have indicated a significant shift in their approach to artificial intelligence spending. According to a survey conducted by Bloomberg Intelligence, these executives plan to allocate nearly half of their AI budget—46%—to domestic products over the next year, a notable increase from the current allocation of 30%. This change reflects a growing trend among Chinese companies to prioritize local suppliers in the face of rising costs associated with AI-related projects. The survey included responses from 60 executives across various sectors, including software, finance, manufacturing, and retail, highlighting a broad consensus on the need for domestic solutions.
The report also revealed that 80% of these executives are experiencing budget overruns in their total infrastructure spending this year, primarily due to the high costs associated with AI initiatives. Major players in China's AI infrastructure, such as Tencent Holdings Ltd., Alibaba Group Holding Ltd., and Huawei Technologies Co., are well-positioned to benefit from this shift towards domestic products. Additionally, AI accelerators from companies like Hygon Information Technology Co. and Cambricon Technologies Corp. are being evaluated by a significant number of survey respondents, indicating a strong interest in local alternatives to foreign technology.
Despite the ongoing popularity of Nvidia's products, the survey suggests that the company's market share may decline as its H20 chips become less accessible due to government pressure on tech firms to avoid using foreign technology. This trend aligns with China's broader strategy to develop its own semiconductor capabilities and reduce reliance on foreign suppliers. The government is investing approximately 2 trillion yuan (around $294 billion) over the next five years to build data centers nationwide, aiming to enhance the use of AI across various sectors, including healthcare and urban management.
However, the growth of Chinese AI firms may face challenges due to a global memory chip shortage, which could hinder their ability to scale effectively. Companies like Semiconductor Manufacturing International Corp. and ChangXin Memory Technologies Inc. are expected to benefit from this situation, as the bottlenecks in the industry shift from computing power to securing high-bandwidth memory chips necessary for rapid data transfer. Overall, the survey results underscore a pivotal moment for China's AI landscape, as domestic companies increasingly take center stage in the development and deployment of AI technologies.
