Gartner forecasts $1.3T semiconductor
- Gartner said on April 8 that worldwide semiconductor revenue should hit $1.32 trillion in 2026, after a $805.3 billion 2025, driven by AI buildouts. - The eye-popping detail is memory: Gartner sees that segment jumping from $216.3 billion to $633.3 billion, with DRAM prices up 125% and NAND 234%. - The catch is broad pain elsewhere — Gartner says memflation could delay non-AI demand into 2028.
Semiconductors are having one of those moments where one part of the market gets so hot it distorts everything around it. That’s basically Gartner’s new call. On April 8, the firm said global chip revenue should reach $1.32 trillion in 2026, up 64% from an expected $805.3 billion in 2025. That would make 2026 the industry’s fastest growth year in roughly two decades — and the reason is not “all chips are booming.” It’s AI infrastructure, plus a memory market that looks almost violently tight. (gartner.com) ### Why is this forecast such a big jump? Because Gartner is no longer describing a normal semiconductor cycle. It’s describing an AI-led squeeze. Hyperscalers are still pouring money into AI data centers, and Gartner says their AI infras(gartner.com)specially memory. AI semiconductors alone are expected to make up about 30% of all chip revenue in 2026. (gartner.com) ### Why does memory matter so much? Because memory is where the forecast gets extreme. Gartner sees memory revenue tripling from $216.3 billion in 2025 to $633.3 billion in 2026. Non-memory grows too, but nowhere near that pace — from $58(gartner.com)in a single year. (gartner.com) ### What’s actually blowing up inside memory? Prices. Gartner says average 2026 prices for DRAM and NAND flash should rise 125% and 234%, with meaningful relief not expected until late 2027. That’s why this isn’t just a unit-demand story. (gartner.com)ounds silly until you look at the numbers. (gartner.com) ### Why would AI push memory this hard? AI servers are memory-hungry in a very specific way. Training and inference clusters need huge amounts of fast memory close to the processor, plus more storage behind it. In 2025, Gartner said high-(gartner.com)n acceleration of a market that was already tilting hard toward AI hardware. (gartner.com) ### Who benefits most from this? The companies already sitting on the AI stack. Gartner’s January ranking showed NVIDIA at $125.7 billion in 2025 semiconductor revenue, far ahead of Samsung at $72.5 billion and SK hynix at $60.6 billion. That ranking matters because it shows where the money is concentrating — AI compute at the top, and memory suppliers right behind it. (gartner.com) ### Who gets squeezed? Regular device markets. Gartner warned in February that soaring memory and SSD costs could push 2026 PC shipments down 10.4% and smartphone shipments down 8.4%. It also projected PC prices up 17% and smartphone prices up 13% versus 2025. So the same shortage dynamics that juice semiconductor revenue can make finished devices more expensive and delay upgrade cycles. (gartner.com) ### Why does this matter beyond chip stocks? Because it reframes the bottleneck. The problem is no longer just “can AI demand stay high?” The problem is whether the supply chain can feed that demand without breaking pricing for everything else. When(gartner.com)ry. (gartner.com) ### Bottom line Gartner’s trillion-dollar forecast is really an AI-and-memory forecast. That’s the important translation. If AI buildouts keep running and memory stays tight into late 2027, chip revenue can surge even while parts of the broader electronics market feel worse, not better. (gartner.com)
