The impact of AI(artificial intelligence) on the semiconductor industry has gradually emerged from the two indicators of sales opportunities and production mode upgrading, including OS manufacturers, EDA, IP manufacturers, IC chip manufacturers have launched a new generation of architecture and product planning for AI applications in 2017, and the impact of AI will continue to expand in 2018. It is expected that the compound annual growth rate of semiconductors from 2018 to 2022 will be 3.1%, and AI will play the main growth momentum of semiconductors.

       Topology Research Research manager Lin Jianhong pointed out that AI is affecting the semiconductor industry from two different ways, one is the sales opportunity, including new applications to bring new products and new technologies, such as more sensors, math accelerators, storage units and communication capabilities, implementation of services, building communication backbone, and synchronous upgrade of data centers and servers. On the other hand, it will bring the upgrading of the production mode of the semiconductor industry.

           From the sales point of view, the key growth brought by AI in 2018 includes the increase in the number of semiconductors carried by a single product, the increase in the average price of major semiconductor products, and the stable volume of new application terminals.

        From the perspective of application, more and more sensors and control components are introduced into vehicles, electric vehicles or advanced driver assistance systems; Voice assistants bring new smart home usage scenarios and product requirements.

         In addition, smart phones introduce a variety of biometric solutions, the demand for data computing, storage and transmission is getting higher and higher, but also promote the chip upgrade demand, the most obvious, including the top three smart phone brands, five high-end mobile phone chip suppliers are providing and using AI acceleration features of the IC and application suite.

          On the other hand, whether from the perspective of AI introduction or Industry 4.0, new production models are reshaping the definition of effective capacity for semiconductor companies. It is expected that in 2018, the difference in the degree of digitalization and the amount of control data will have an increasingly significant impact on various manufacturers, and semiconductor manufacturers in various regions will have different degrees of pressure and opportunities due to the differences in the mastery of new production modes by various companies.