AI Chip Market and Development Trend Analysis Report: Jensen Huang Claims AI May Surpass Human Intelligence Within 5 Years, Developing Compliant Chips for Chinese Market
-
On November 29 local time, NVIDIA CEO Jensen Huang stated at the DealBook Summit that if artificial general intelligence (AGI) is defined as computers or software that are 'competitive' with human intelligence in tests, then 'AI could pass these tests within the next five years.'
During the event on the 29th, Huang explained how NVIDIA's products rely on numerous components from around the world and stated: 'We still need 10 to 20 years to achieve supply chain independence. This is not something that can be accomplished in a year or two.' Huang also publicly stated that China is the largest chip market and NVIDIA is developing products for China that will not face restrictions: 'We must produce new chips that comply with regulations. Once they meet the requirements, we will return to China.'
NVIDIA and Intel Plan to Launch Modified AI Chips for Chinese Market, NVIDIA Products May Enter Mass Production by Year-End
Multiple industry sources revealed that NVIDIA is developing modified chips for the Chinese market, reducing performance to meet new regulatory parameters. One source mentioned that discussions with NVIDIA are ongoing, with mass production expected by year-end. Meanwhile, Intel is also planning to release a modified version of Gaudi 2. Currently, neither NVIDIA nor Intel has officially responded.
AI chips, as the computational foundation for large models and AI applications, are becoming increasingly critical. AI chips, also known as AI accelerators or compute cards, are specialized modules for handling 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. Much of AI's data processing involves matrix multiplication and addition. GPUs, with their massive parallel processing capabilities, offer a cost-effective solution but consume more power. FPGAs, equipped with built-in DSP modules and local memory, are more energy-efficient but typically more expensive.
From a technical perspective, the first 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 like Intel, Google, NVIDIA, Qualcomm, and IBM, at least one CPU is required to control these systems. However, when processing streaming data in parallel, various types of coprocessors become necessary.
According to the '2023-2028 China AI Chip Market and Development Trend Analysis Report' published by Zhongyan Puhua Industrial Research Institute, statistical analysis shows:
With the continuous development of artificial intelligence technology and the expansion of application scenarios, the demand for AI chips will continue to grow. In the future, China's top ten AI chip companies will focus more on advanced technologies such as deep learning and edge computing, developing more powerful and highly integrated AI chips to meet the growing demands of various industries.
The upstream of the AI chip industry chain is primarily chip design, which can be further subdivided into three categories based on business models: IP design, chip design foundry, and chip design. Most companies in this sector are IC design companies.
AI chips serve as the "brain" of artificial intelligence. The main types of AI chips currently include CPUs, GPUs (Graphics Processing Units), FPGAs (Field-Programmable Gate Arrays), DSPs, ASICs (Application-Specific Integrated Circuits designed for neural network algorithms), and brain-inspired chips. ASICs are expected to replace general-purpose chips as the mainstay of AI chips in the coming years.
The downstream of the AI chip industry chain mainly consists of system integrators and application enterprises, such as AI solution providers. The most popular application areas include autonomous driving, smartphones, robotics, and security.
AI chips are the core of the information industry and a key force driving the new wave of technological revolution and industrial transformation. The AI chip industry can be divided into three competitive tiers based on companies' registered capital. Companies with registered capital exceeding 10 billion yuan include Huawei and Horizon Robotics; those with registered capital between 1-10 billion yuan include Cambricon, AISpeech, and Enflame; and the remaining companies have registered capital below 1 billion yuan.
The AI chip industry is capital-intensive and technology-intensive, with high technical barriers. As a result, China's AI chip companies are concentrated in economically developed and talent-rich regions such as Beijing, Shenzhen, and Shanghai.
AI Chip Industry Landscape
In the AI chip sector, foreign chip giants dominate most of the market share. Globally, major players in AI chip development include Intel, NVIDIA, Qualcomm, and Google.
Faced with the monopolistic pressure from foreign manufacturers, domestic internet companies and AI chip startups are actively entering the market.
Driven by both policy support and market demand, Chinese AI chip manufacturers are striving to catch up.
Domestic AI chips, represented by Cambricon's Siyuan series and Huawei's Ascend series, have reached high performance levels and are expected to accelerate domestic substitution and enter a period of rapid development.
Under the AI wave, industries such as cloud computing, smart vehicles, and intelligent robots are rapidly developing, increasing the demand for AI chips. The AI chip market size is expected to continue growing.
Expected to expand to more than 10 times the current size in the next 10 years
With the expansion of deep learning and AI deployment, the demand for AI chips is rising. It is expected that the market size will expand to more than 10 times the current size in the next 10 years.
It is predicted that by 2025, China's AI chip market size will reach 178 billion yuan, nearly doubling from 2022. The CAGR of China's AI chip market size from 2021 to 2025 is 42.9%, faster than the global market growth rate (32.1%) during the same period.
With the continuous popularity of large AI models, an unprecedented battle for self-developed AI chips has begun. Recently, OpenAI, the developer of ChatGPT, announced that it will start developing its own AI chips. Similarly, Microsoft plans to launch its first AI chip at its annual developer conference next month. Besides OpenAI and Microsoft, many other manufacturers are also gearing up, with world-leading tech companies like Intel, Google, Amazon, and Meta developing their own AI chips.
Analysis of China's AI Chip Market Size in 2023
With global competition intensifying, China's AI chip industry is facing a critical opportunity period. Data shows that by 2023, China's AI chip market size will further expand to 120.6 billion yuan.