AI chip

Driven by adoption of artificial intelligence (AI) across multiple industries, the global AI chip market generated nearly $29 billion in 2022, according to a report from Next Move Strategy Consulting.

The market is expected to generate nearly $305 billion by the end of the decade, boasting a compound annual growth rate (CAGR) of 29% from 2023 to 2030, the report projected.

These types of chips are a type of semiconductor designed specifically for AI applications and are more efficient and faster than traditional chips.

The broader chip market includes a wide variety of semiconductors, including CPUs, GPUs, memory chips and networking chips. This makes them ideal for a variety of applications, such as self-driving cars, facial recognition and natural language processing (NLP).

AI chips are typically in the form of a system-on-a-chip (SoC), which integrates all or most components of a computer or other electronic system on a single chip.

“This makes them smaller, more powerful and more energy-efficient than traditional chips,” explains Next Move analyst Shyam Gupta. “AI chips are a relatively new segment of the chip market, but they are growing rapidly due to the increasing demand for AI-powered applications.”

Gupta adds the AI chip industry is growing rapidly due to the increasing demand for autonomous and semi-autonomous driving.

“AI chips are used in these vehicles to interpret the surroundings, make intelligent choices and navigate the roads independently,” he says.

The capacity of AI to handle extensive data and process it, learn from past encounters and adjust to diverse scenarios is pivotal in attaining safe and efficient autonomous driving. For example, FABU’s Phoenix-100 perception chip is specifically engineered for autonomous driving solutions, offering the real-time and highly accurate perception of the surrounding environment.

Moreover, the industry for AI chips is further thriving due to the rapid automation in warehouses, where robots and other machines are extensively used to automate tasks and improve accuracy.

“AI chips are used to control these robots and machines, as well as process data received from different sensors,” Gupta notes. “These data are used to make decisions about how to automate tasks in warehouses.”

John Leddy, managing director, technology division, JLL, notes as the AI market continues to expand and grow its capabilities, demand for semiconductors will match the AI industry growth.

“This will require semiconductor designers and manufacturers to develop new, more efficient chips to process the increase in demands on computing, storage and memory,” he says. “This will require expansion and continued investment in manufacturing capabilities.”

He adds that despite the recent decline in semiconductor demand due to slumping PC production post pandemic, JLL expects to see the semiconductor industry grow at a 6-8% annual rate, driven by automotive, data storage and wireless industries.

“Demand for semiconductors specifically for AI applications may increase and accelerate this annual growth with AI market forecasts calling for a ten to twentyfold growth by 2030,” Leddy says.

Gupta points to IBM, AMD, Nvidia, Intel and Qualcomm as the leading companies in the AI chip industry, which is expected to see a significant evolution in the coming years due to the introduction of 2-nanometer (nm) chipsets in smartphones.

With the integration of 2nm chipsets into smartphones, manufacturers can significantly enhance energy efficiency by reducing power consumption, increasing performance per watt, optimizing AI acceleration, incorporating advanced power management features and improving thermal efficiency.

“These advancements contribute to prolonged battery life and better overall energy utilization, providing users with longer lasting and more efficient smartphones,” Gupta says.

Despite the enormous growth potential, he cautions growing intellectual property (IP) and patent disputes are obstructing the development and distribution of AI chips and constraining market growth.

“Patent disputes can be costly and time-consuming to resolve, and they can also obstruct the development and distribution of AI chips,” Gupta explains. “It can discourage companies from investing in R&D.”

For example, in 2020, Nvidia sued rival chipmaker Intel for patent infringement. The lawsuit alleged that Intel’s AI chips infringed on Nvidia’s patents for GPU technology. The lawsuit was settled in 2022, with Intel agreeing to pay Nvidia a licensing fee.

In addition, Gupta notes rising geopolitical tensions that give rise to export regulations imposed by governments on the AI chip industry are likely to restrict the growth of the market.

“For example, the US Department of Commerce placed new restrictions on the export of AI chips used in high-performance computing applications, especially to China,” he says. “This regulation will impact the cost of AI chips in the coming years.”