According to Gartner, the market for integrated circuits (ICs) designed for AI workloads is projected to reach $53.4 billion this year, marking a substantial 20.9% increase compared to the previous year.
The expansion in generative AI and the growing utilization of diverse AI-driven applications across data centers, edge infrastructure, and endpoint devices are necessitating the adoption of high-performance GPUs and finely-tuned semiconductor devices. Alan Priestley, Vice President at Gartner, explains that this trend is fueling the development and deployment of AI chips.
Forecasts indicate that the revenue from AI semiconductors will experience a further 25.6% surge in 2024, reaching $67.1 billion (refer to Table 1). Looking ahead, it is projected that AI chip revenue will soar to $119.4 billion by 2027.
Table 1. AI Semiconductors Revenue Forecast, Worldwide, 2022-2024 (Millions of U.S. Dollars)
| Year | 2022 | 2023 | 2024 |
|---|---|---|---|
| Revenue ($M) | 44,220 | 53,445 | 67,148 |
Within the consumer electronics domain, Gartner analysts estimate that the value of AI-enabled application processors incorporated into devices will climb to $1.2 billion by the close of 2023, up from $558 million in 2022.
Priestley emphasizes that, for numerous organizations, extensive deployments of tailor-made AI chips will replace the existing predominant chip architecture – discrete GPUs – particularly for various AI-based workloads, especially those grounded in generative AI techniques.
The upsurge in generative AI is also fostering the demand for high-performance computing systems, both for development and deployment. Several vendors are offering systems centered around high-performance GPUs and networking equipment, reaping notable short-term advantages.
In the longer term, as hyperscale entities seek efficient and cost-effective means of deploying these applications, Gartner foresees an escalation in their adoption of custom-designed AI chips.