• Memflation–Gartner’s term for AI-driven memory price inflation–is pushing semiconductor revenue to a 20-year high, while quietly freezing non-AI demand until 2028.
• The US$1.3 trillion headline is the good news. The warning buried underneath it is what enterprise buyers should actually be reading.

The semiconductor industry is having its biggest year in two decades. Global revenue is projected to exceed US$1.3 trillion in 2026, growing 64% year on year, according to Gartner’s latest forecast published on April 8. That figure has generated predictable headlines about the AI boom’s reach into hardware markets. What most of those headlines skipped is the second half of the report.

Gartner’s senior principal analyst Rajeev Rajput coined a term for what is actually driving those numbers: memflation. The phenomenon describes what happens when demand for AI infrastructure consumes such a disproportionate share of memory production that prices across the entire stack–DRAM, NAND flash, and downstream components–spike sharply for everyone else. 

Memory revenue is expected to nearly triple this year, rising from US$216.3 billion in 2025 to US$633.3 billion in 2026. DRAM prices are forecast to rise 125%, and NAND flash prices by 234%. No meaningful pricing relief is expected before late 2027.

That is not a footnote. That is the structural reality underneath the record revenue figure.

Memflation is a supply diversion problem, not a demand boom

The distinction matters. Global semiconductor revenue is growing not because demand has broadly expanded across every end market, but because AI infrastructure spending is consuming available production and inflating prices across the board.

Rajput was direct about the consequence: “Memflation will destroy, or at least delay, non-AI demand into 2028, to varying degrees depending on the application.” The AI wave is not lifting all boats. It is redirecting the flow.

Hyperscaler investment in AI infrastructure is expected to increase by more than 50% in 2026, driving demand for GPUs and custom chips. AI semiconductors will account for roughly 30% of total semiconductor revenue this year. The remaining 70% of the market–consumer devices, industrial systems, automotive, enterprise servers not tied to AI workloads–is competing for a shrinking share of available supply at inflated prices.

The pressure on the supply chain is already visible downstream. According to analyst firm Omdia, more than 70% of channel partners reported price increases in 2026, with over 90% experiencing shipment delays. PC and server vendors have been passing those added costs through to buyers.

NVIDIA’s position in all of this is worth noting. According to a separate Gartner analysis from January, Nvidia surpassed Samsung as the largest semiconductor provider by revenue for the first time in late 2025, as AI processors surpassed US$200 billion in sales. 

The single most important company in the semiconductor market now derives most of its revenue from a product category that represents 30% of industry demand–and that concentration is precisely what is distorting conditions for everyone else.

What Gartner actually wants CIOs to do with this information

The headline figure tends to get cited in boardrooms as validation that the AI infrastructure investment cycle is durable. Gartner’s analyst is not saying otherwise, but the practical advice that came with the forecast is considerably more cautious.

Rajput advised that “CIOs and IT leaders should be cautious about signing supply agreements with unfavourable pricing terms that extend beyond 2027.” That is a specific, time-stamped warning about procurement decisions being made right now, under current market conditions, that will bind enterprises to pricing structures that may look very different once memflation moderates.

Gartner expects the market to grow a further 17% in 2027 to nearly US$1.6 trillion as demand begins to level off. Pricing pressure is expected to moderate–not disappear–through the second half of 2026, before more meaningful relief arrives later. Any enterprise that locks in multi-year supply agreements at 2026 peak pricing is taking a bet that relief does not arrive on the schedule Gartner is projecting.

Micron Technology’s own CEO, Sanjay Mehrotra, made the supply dynamic clear during a Q2 2026 earnings call. “Data centres are becoming a bigger and bigger part of the industry,” Mehrotra said. “So of course, a bigger portion of the supply goes there, and that’s the main driver of growth for the industry as well as for Micron itself.” 

The largest memory manufacturer in the US is openly acknowledging that data centre demand is crowding out other markets. Enterprise buyers should take that at face value.

Memflation’s shelf life

Gartner is explicit that the memory price surge is “profound, but not perennial.” The firm does not expect DRAM and NAND flash prices to sustain their 2026 trajectory indefinitely. The correction will come; the question is timing, and that timing has direct implications for any hardware procurement cycle running through 2027 and beyond.

The semiconductor industry’s record revenue year, in other words, contains within it the clearest possible signal about when conditions will shift. The CIOs who read past the US$1.3 trillion headline will be better positioned than those who used it as justification for locking in long-term pricing.

Source: Gartner Forecasts Worldwide Semiconductor Revenue to Exceed $1.3 Trillion in 2026, published April 8, 2026. Additional data from Channel Dive/Omdia and Gartner January 2026 semiconductor vendor ranking.