Analysis of AI Chip Market Development: China's AI Chip Shipments Expected to Reach 1.335 Million Units in 2023
-
Under the policy guidance of moderately advancing the construction of new infrastructure, investments in AI computing power infrastructure have increased. Internet companies, telecom operators, and governments at all levels are actively participating in the construction of intelligent computing centers. According to incomplete statistics, as of August 2023, over 30 cities across China have built intelligent computing centers.
Driven by the wave of large AI models, the market scale for AI chips, servers, and data centers is expected to grow significantly. IDC predicts that China's AI chip shipments will reach 1.335 million units in 2023, a year-on-year increase of 22.5%.
Regarding AI servers, IDC forecasts that the global AI hardware market (servers) will grow from $19.5 billion in 2022 to $34.7 billion in 2026, with a compound annual growth rate of 17.3%.
AI Chip Market Development Analysis
AI chips, also known as AI accelerators or computing cards, are modules specifically designed to handle large-scale computational tasks in AI applications (non-computational tasks are still managed by CPUs). Currently, AI chips are mainly categorized into GPUs, FPGAs, and ASICs.
In terms of technological approaches, the first generation of AI chips in the market included various combinations of off-the-shelf CPUs, GPUs, FPGAs, and DSPs. While new designs are being developed by companies such as Intel, Google, NVIDIA, Qualcomm, and IBM, at least one CPU is still required to control these systems. However, various types of coprocessors become necessary when processing parallelized streaming data.
In recent years, the Chinese government has introduced multiple policies to encourage the development and innovation of the AI chip industry, including the National New Generation Artificial Intelligence Development Plan and the Notice on Deepening 'Mass Entrepreneurship and Innovation' Policies to Promote Public Entrepreneurship and Innovation.
Key cities such as Beijing, Shanghai, and Shenzhen have successively released policies to support the rapid development of the AI industry, including breakthroughs in AI chip innovation, strengthening R&D in AI chips and smart sensors, and accelerating the construction of intelligent computing power.
On December 6, local time, AMD announced at its 'Advancing AI' event in San Jose, USA, the launch of its new AI chip series, Instinct MI300, including models such as MI300A and MI300X, along with corresponding software and services. The MI300A has already entered mass production, while the MI300X has begun shipping.
It is reported that the MI300X adopts AMD's new CDNA 3 architecture, featuring 153 billion transistors. It combines 5nm and 6nm process chiplets, utilizes 3D packaging technology, and is equipped with 192GB of HBM3 memory.
On August 29, Google also unveiled its new AI chip at the annual Google Cloud Next conference in San Francisco—the fifth-generation custom Tensor Processing Unit (TPU) chip, TPU v5e, designed for large model training and inference. Compared to the previous generation, the TPU v5e offers 2x training performance per dollar and 2.5x inference performance per dollar.
On November 15, Microsoft introduced the Maia chip at its Ignite developer conference in Seattle, aimed at accelerating AI computing tasks. The Maia chip is designed to run large language models and differs from Nvidia's networking technology by connecting via standard Ethernet cables.
AI Chip Market Opportunities
It is widely recognized that developing proprietary AI chips is an inevitable trend as industries mature. Any company experiencing a significant increase in AI computing demand will need its own chips to achieve optimal performance.
In the context of the digital economy, given the unprecedented market potential in AI, building an in-house chip supply chain has become a strategic priority for global tech giants. Companies like Google, Microsoft, and Amazon—software and cloud service leaders—are naturally entering the chip competition.
Global competition is intensifying, and China's AI chip industry is facing a critical opportunity period. Data shows that China's AI chip market reached 42.7 billion yuan in 2021, a 124% year-on-year increase. In 2022, the market doubled again to 85 billion yuan. China Post Securities predicts that by 2023, China's AI chip market will expand further to 120.6 billion yuan.
In this fiercely competitive market, timely and effective decision-making is key to success. Reports from China Research Network provide detailed analysis of China's AI chip industry, covering development status, competition, and supply-demand dynamics, along with insights into policy, economic, social, and technological factors. These reports reveal potential market demands and opportunities, offering strategic investors and corporate leaders accurate market intelligence for planning. They also serve as valuable references for government agencies.