Gartner forecast on 24 August that worldwide semiconductor revenue will reach $1.6 trillion in 2026, up 92% from $809 billion in 2025. Growth on that scale is not normal for this industry, which historically moves in cycles of tens of percent, not doublings.
The concentration behind the number is as striking as its size. Gartner director analyst Ben Lee put the AI data centre ecosystem at 36.5% of all semiconductor revenue in 2026, rising to more than 53% by 2030. On that trajectory, most of the semiconductor industry by value ends up serving a single application category.
Lee also made the point that the effect is not confined to accelerators. AI build-out pulls through CPUs, networking silicon, power management, analogue parts and optical interconnect — which is why component categories with no obvious connection to machine learning have been repricing too.
For a small product company the read-across is straightforward and slightly uncomfortable. Priority at foundries, packaging houses and memory suppliers follows volume, and consumer or industrial designs in the thousands do not have much of it. Planning for longer lead times and holding a little more buffer stock on critical parts is cheap insurance compared with a line stoppage.
Source: Gartner