This article first appeared in The Edge Malaysia Weekly on March 26, 2018 - April 1, 2018
THE following are more thoughts from the four experts.
The Edge: It is anticipated that AI will be the next growth driver for the semiconductor industry. What is your view on that?
Priestley: Since 2016, AI has become the most impactful and disruptive technology to all technology and service providers. Gartner’s view is that the market for AI systems is still evolving. [This] is creating an opportunity for new customised and semi-customised semiconductor devices.
Armstrong: AI is the next big growth driver for a lot of industries. Thanks to automation, AI will drive a new wave of efficiencies and cost-savings, among other things.
Heng: Many big corporations are investing quite substantially to acquire and to have a share in AI. Coupled with the release of 5G, I think it will grow.
Chu: AI is definitely good for us because big data analytics requires a lot of servers. A server is an expensive electronic board that requires a lot of inspection and quality control, and that’s where we come in.
What are the key drivers of demand growth in AI semiconductors and, more importantly, how will AI adoption impact industry dynamics?
Priestley: The driving force behind semiconductors and the electronics industry has shifted significantly during the past few years. The momentum of the internet of things and the growing numbers of IoT projects have attracted the semiconductor industry to invest in technologies to support the growing demands of intelligent products (including connected cars) and servers.
Armstrong: Cost-effectiveness and the desire to remain competitive (or ahead of competitors) will be the key areas that affect demand in this area. AI costs money and those who want it enough will make it happen.
Heng: If AI takes off successfully, semiconductors will see another bullish cycle and opportunity.
Chu: AI means a lot of computational power, you need a lot of data centres. Each data centre requires a lot of switchers, routers and other hardware infrastructure, which need reliable boards that need to go through vision inspection. AI needs to have a strong connection between the sensors to collect information in order to make the right decisions. This will drive the demand for semiconductors.
The automotive segment is expected to be the next big wave for the semiconductor industry. The most common AI application is for autonomous vehicles. If we draw a line between AI and automotive, which is more likely to be the next growth driver for the global semiconductor industry, AI or automotive?
Priestley: Automotive electronics was one of the fastest-growing segments in the semiconductor industry in 2016, as sales of semiconductor products to the automotive industry were fuelled by increasing motor vehicle production and increasing electronic content. The automotive sector continues to be a bright spot in the semiconductor market.
Armstrong: AI is likely going to drive more innovation and opportunities but both (AI and automotive) will have a significant impact on the future of semiconductors.
Heng: I think there is no need to distinguish between that because AI will generate a large set of new technologies — that in itself is driving growth. We are talking about technology from wafer fabrication, integration, assembly and testing. As for hardware, it will be mainly driven by existing and new sensors. Each AI system will require the integration of many sensors, which will support quantum growth in semiconductor quantity, and drive capital expenditure and new material too.
Chu: AI covers such a huge area. We have smart city, smart car, smart building, smart toilet, smartphone — all these things are related to AI. Huawei has launched its AI chipset, which needs processors and sensors, thus driving the demand for chips.
Save by subscribing to us for your print and/or digital copy.
P/S: The Edge is also available on Apple's AppStore and Androids' Google Play.