Beijing's AI chip self-reliance push with local subsidies

Beijing's AI chip self-reliance push with local subsidies

2024-04-27 semicon

Beijing offers subsidies for domestic AI chip purchases to bolster tech autonomy and counter US export restrictions.

The Semiconductor Industry at a Crossroads

In the wake of escalating trade tensions and the imposition of US export controls, Beijing’s strategic move to offer subsidies for domestic AI chip purchases is a clear signal of its commitment to the semiconductor industry, specifically in the realm of artificial intelligence (AI) computing. The Beijing Municipal Bureau of Economy and Information Technology’s decision to provide financial incentives for the procurement of China-made graphics processing units (GPUs) is a pivotal step towards achieving technological self-sufficiency[2].

Benefits and Operational Mechanism

The initiative to support local firms in acquiring domestically produced AI chips is expected to bolster Beijing’s computing infrastructure and foster an environment conducive to the growth of intelligent computing services. By leveraging subsidies, the city aims to accelerate the adoption of AI processors, which are paramount for the training of large language models and the provision of generative AI services akin to ChatGPT[1]. The development of core technologies such as AI processors, operating systems, and databases within the city’s jurisdiction aligns with the broader goal of reducing dependence on foreign technology amid stringent US sanctions on chip exports to China[1].

Silicon Photonics and Quantum Computing: A Technological Leap

The Beijing government’s focus is not limited to traditional semiconductor technology; it is also targeting advancements in silicon photonics and quantum computing chips. These technologies represent the cutting edge of the semiconductor industry, and Beijing’s interest in them underscores the city’s ambition to be at the forefront of the next wave of technological breakthroughs. Silicon photonics involves the use of photonic systems to transport data among computer chips, which promises faster and more energy-efficient data processing. Quantum computing, on the other hand, harnesses the principles of quantum mechanics to perform calculations at speeds unattainable by traditional computers. These innovations could have far-reaching implications for various industries, including cryptography, drug discovery, and complex system modeling[1].

Achieving Autonomy by 2027

The subsidies are part of a broader strategy by Beijing to reach full self-reliance in smart computing infrastructure by 2027. The Ascend 910 chips from Huawei Technologies have been highlighted as a potential domestic alternative to foreign chips, with intelligent computing centers established by the government leading the purchase of these domestic AI chips[3]. This move is a direct response to the U.S. Department of Commerce’s export restrictions, which have targeted advanced chips like Nvidia’s A800, H800, and H100, all critical components for AI applications[3].

Bronnen


AI chips Beijing