AI Infrastructure Keeps Global IT Spending on Track for a $6.31 Trillion Year

Updated: 2026-04-24

AI spending is reshaping the IT market mix

Global IT spending is forecast to reach $6.31 trillion in 2026, according to Gartner, with AI infrastructure and software adoption driving much of the acceleration. The projected year-on-year increase is notable not just for its size, but for the way it concentrates value in data centers, memory, compute platforms, and the software layers built around generative AI.

The strongest growth is expected in data-center systems, where spending is set to climb by more than half. That reflects the scale of hyperscale AI buildouts now underway. As cloud and enterprise operators expand training and inference capacity, they are pulling through demand for advanced processors, accelerators, interconnect, and high-bandwidth memory.

For electronics and semiconductor companies, that mix matters. The forecast points to a market where AI-oriented hardware is capturing disproportionate investment, while memory suppliers benefit from both rising demand and tighter supply. Gartner’s view implies that pricing power in advanced memory could remain firm as long as AI deployments continue to widen.

Software is also expected to post another strong year, helped by continued investment in generative AI model development and deployment. That keeps software near the center of enterprise AI budgets even as hardware spending surges, reinforcing a two-sided growth cycle in which platforms and infrastructure rise together.

By contrast, device spending is projected to grow at a more modest pace. Higher memory costs are pushing up average selling prices and slowing replacement cycles, particularly in categories that are already margin-sensitive. In other words, AI is lifting the overall market, but not every segment benefits at the same speed.

For European designers, distributors, and component suppliers, the forecast is a useful signal: the next wave of demand is likely to cluster around AI-capable infrastructure, memory-intensive designs, and the services needed to operationalize them. The addressable market is getting larger, but it is also becoming more uneven and more infrastructure-led.