# AI infrastructure spending hits $2.59 trillion

> Gartner's new forecast shows hardware makers are capturing most of the revenue as compute and memory costs surge.
- Title: AI infrastructure spending hits $2.59 trillion · Cloud Wars
- Summary: Gartner's new forecast shows hardware makers are capturing most of the revenue as compute and memory costs surge. The AI boom's profits are flowing to…
- Keywords: gartner, ai infrastructure, capex spending, nvidia, hardware costs, dram prices, compute, memory, hype cycle, return on investment, infrastructure, spending
- Source: Cloud Wars — https://cloudwars.com/cloud-wars-ai-minute/ai-infrastructure-spending-hits-2-59-trillion-as-hardware-costs-reshape-the-market
- Author: Dewain Robinson
- Published: 2026-09-18T12:00:00+00:00
- Read time: 1 min
- Topics: gartner, ai infrastructure, capex spending, nvidia, hardware costs, dram prices, compute, memory, hype cycle, return on investment
## AI infrastructure hits $2.59 trillion
Gartner reports companies are investing $2.59 trillion in AI infrastructure capex. Cash flow is struggling to keep pace with these massive expenditures.
## Nvidia revenue up 117 percent
Nvidia hit $89 billion in revenue in one quarter, up 117% year over year. Demand far exceeds their ability to produce equipment.
## Memory prices surge 400 percent
DRAM prices have jumped 400% since early 2024. Combined with compute costs, hardware manufacturers are capturing most AI revenue.
## Trough of disillusionment coming in 2026
Gartner predicts the AI hype cycle will hit its trough in 2026. The market will shift focus to actual return on investment.
## Demand may normalize after hype fades
As the hype cycle cools, demand on compute and memory infrastructure could normalize. This may open up new opportunities in the market.
## Key takeaway

The AI boom's profits are flowing to hardware manufacturers, not software companies, as compute and memory costs skyrocket.