Chips that are used for artificial intelligence (AI) applications are set for massive growth in the next five years as use of the technology expands across a variety of markets, according to new research from IHS Markit.
Memory and processing semiconductors with AI functions will grow to $128.9 billion in 2025, three times the $42.8 billion total in 2019. In the AI segment, worldwide revenue from memory devices in AI applications will increase to $60.4 billion in 2025, up from $20.6 billion in 2019. The process segment will grow quicker to $68.5 billion in 2025, up from $22.2 billion last year, IHS Markit said.
AI chips are used in numerous markets including automotive, communication, computers, consumer electronics, industrial and healthcare. The largest market for memory devices in AI applications is the computer segment while the communication, consumer electronics, industrial and healthcare sectors will growth faster but will not be worth as much in the long term.
“Semiconductors represent the foundation of the AI supply chain, providing the essential processing and memory capabilities required for every artificial intelligence application on earth,” said Luca De Ambroggi, senior research director for AI at IHS Markit. “AI is already propelling massive demand growth for microchips. However, the technology also is changing the shape of the chip market, redefining traditional processor architectures and memory interfaces to suit new performance demands.”
The market research firm said that startups are looking to offer completely new architectures that will challenge the market of traditional devices used for AI processing such as graphics processing units (GPUs), field programmable gate arrays (FPGAs), microprocessor units (MPUs), microcontroller units (MCUs) and digital signal processors (DSPs). These new architectures include capabilities such as integrated vector-processing that helps accelerate deep-learning tasks.
Because of these new architectures, the old definitions of what makes an MPU, DSP or MCU will blur together into in the AI era.
Additionally, the introduction of AI-related capabilities into various devices means that these traditional processor classes are evolving to the point where they are no longer recognizable as distinct categories.
“Increasingly, designers of AI-enabled systems are using highly integrated heterogenous processing solutions, such application-specific integrated circuits (ASICs) and system-on-chip (SoC) solutions,” De Ambroggi said. “With processor makers offering turnkey, heterogenous processing solutions using these ASICs and SOCs, it makes less difference to system designers whether their AI algorithm is executed on a GPU, CPU or DSP.”
