Worldwide semiconductor revenue is forecast to reach about $1.6 trillion in 2026 as the AI infrastructure buildout creates exceptional demand for memory and other data center components, according to Gartner.
The research firm expects semiconductor revenue to increase 92% in 2026, from $809 billion in 2025. Gartner forecasts another sizable gain in 2027, when worldwide revenue is projected to approach $1.94 trillion.
Driving much of this expansion is memory, which Gartner expects to generate $837.3 billion in 2026, compared with $220.1 billion last year. By 2027, memory revenue is forecast to reach nearly $1.08 trillion.
The economics of the semiconductor sector are clearly changing. Memory represented 27% of total industry revenue in 2025 but is forecast to account for 54% this year. NAND flash revenue is projected to rise 371.9% in 2026, while DRAM revenue is expected to grow 246.6%.
The remarkable demand for memory stems in large part from the scale of current AI systems, which require large amounts of memory, including HBM, to feed data rapidly to processors. As companies build larger AI clusters, memory capacity becomes a key component of overall system performance.
For the enterprise sector, this growth creates higher costs and tight supplies. Additional manufacturing capacity is expected to become available in 2027, but Gartner forecasts that demand will remain strong enough to keep the memory market constrained. Significant pricing relief is not expected until late 2027.
Warning About Supply Contracts
The potential for an extended period of elevated prices creates a purchasing challenge for executives planning data center upgrades and AI deployments. Gartner has advised tech leaders to be wary of supply contracts that lock in unfavorable prices beyond 2027, when market conditions could begin to ease.
The semiconductor boom extends well beyond memory. Gartner forecasts nonmemory semiconductor revenue will rise 21.9%, from $589 billion in 2025 to $717.9 billion this year, before reaching $864.4 billion in 2027.
AI clusters require an extensive supporting infrastructure in addition to GPUs and other accelerators. CPUs, networking silicon, optical interconnects, analog components and power management gear all get more expensive as AI systems grow larger and require more electricity and data throughput.
Hyperscalers are of course a key driver of demand. Their spending on AI infrastructure is projected to rise more than 50% in 2026, supporting purchases of GPUs, TPUs and other AI accelerators.
The result is a semiconductor market whose revenue mix is shifting heavily toward AI infrastructure. Gartner expects AI data centers to represent 36.5% of worldwide semiconductor revenue in 2026. By 2030, their share is forecast to exceed 53%.
Bottom line: Ben Lee, Director Analyst at Gartner said that “the semiconductor industry is entering a fundamentally new phase of growth.”
