The AI Wave Has Spawned New Growth Points In The Market
Artificial intelligence (AI) chips are the key infrastructure supporting the development of artificial intelligence technology and industry, and have an important strategic position. At present, artificial intelligence technology and industry continue to develop at a high speed, and a technology industry ecology consisting of chips, data, development frameworks, algorithms, and applications has basically formed. The scale of the artificial intelligence chip industry is growing rapidly. The wave of artificial intelligence development has given birth to new growth points in the chip market. The artificial intelligence chip market maintains strong growth capabilities. According to IDC data, the global overall AI chip market reached US$4 billion in 2017. By 2022, the overall AI chip market size will be Reached 35.2 billion U.S. dollars, and the CAGR was greater than 55%. China’s artificial intelligence chip market is also developing well. According to the “China Artificial Intelligence Chip Industry Development White Paper” released by the research data agency CCID Consulting at the 2019 World Artificial Intelligence Conference, China’s AI chip market continued to maintain rapid growth in 2018. The market scale reached 8.008 billion yuan, a year-on-year increase of 50.2%. Affected by many favorable factors such as the macro policy environment, technological progress and upgrading, and the popularization of artificial intelligence applications, China's AI chip market will further mature. It is estimated that in the next 5 years, the scale of China's chip market will maintain a growth rate of 40%-50%. In 2024, the market will reach 78.5 billion. At present, artificial intelligence chip products mainly cover three areas: cloud training, cloud reasoning, and terminal reasoning. In the field of cloud training, model training needs to achieve high-performance, high-precision, and general computing capabilities. Nvidia has an absolute advantage in the cloud training market with its GPU industry ecology. Google and Intel develop ASIC chips to challenge; In the field of cloud reasoning, energy-efficient GPUs, FPGAs, and ASICs have become the layout schemes of various enterprises. Google, Intel, Xilinx and other manufacturers have laid out their plans, and the future market structure is undecided; In the field of terminal reasoning, chips need to meet the needs of differentiated scenarios such as low latency or low power consumption. There are many market segments. ASIC chips have gradually become the mainstream path. Image recognition and speech recognition are key application areas. Participants include not only Qualcomm, Intel , NVIDIA and other traditional chip giants, as well as new entrants such as Google, HiSilicon, Cambrian, and Horizon. AI chip application innovation is active, and a wider range of applications have been achieved in cloud computing, autonomous driving, smart security, smart phones and other fields, and the future market prospects are ...
Nov 09, 2020