Remarkable growth is predicted from today’s $8 billion market, driven by cloud-based virtual assistants and financial services companies.
The worldwide market for AI accelerator chipsets is predicted to grow from $8 billion in 2019 to reach $70 billion by 2026, according to a new report by analysts Global Market Insights (GMI). This represents a remarkable uptake in the technology from its present position — a CAGR of 35%.
This growth is due in part to the increase in popularity of e-commerce platforms that use AI-based recommendation engines housed in the cloud. The growing acceptance of cloud-powered virtual assistants which use AI to understand speech is also driving demand. Financial services companies are also extensively adopting AI to improve fraud detection and better calculate lending risks.
Increasing use of cloud AI services in practically every industry has reduced operational costs and enhanced efficiency. The cloud had 40% of the market share in 2019 and this will expand at a rate of 30% through 2026, the analysts said.
The European market is expected to experience a surge in demand due to the region’s many companies working on systems for autonomous driving. Autonomous vehicles use AI chipsets to interpret data from cameras and combine it with information from other sensors such as Lidar and ultrasound to build up a picture of their environment. According to the report, the European market will grow 33% between 2020 and 2026 driven by the autonomous vehicle industry in this region.
In Asia-Pacific, the market is expected to grow 40% over the same time frame. China has a healthy ecosystem of young companies and startups, including Cambricon and Horizon Robotics, thanks to a Chinese government initiative begun in 2017. This initiative, the New Generation Artificial Intelligence Development Plan, focuses on investment in the domestic AI industry to the tune of $150 billion by 2030.
The technology leader in the AI accelerator market today is undoubtedly the GPU, with more than 40% share in 2019, according to GMI’s report. GMI also pointed out that several major chip companies have built SoCs which include GPU IP for AI acceleration.
FPGAs, another technology widely used for AI acceleration, are due to grow their share with a CAGR of nearly 25% in the coming years, while the ASIC segment will grow at 45% CAGR.
The report also notes that 25% of the entire AI chipset market in 2019 was used to understand language using NLP (natural language processing) algorithms. This application, used to power smart speakers, automotive infotainment systems and voice controlled appliances and IoT devices, will grow 35% by 2026, the report said.