Updated
Updated · Network World · Aug 12
AI Memory Crunch Extends IT Lead Times to 18 Months as Prices Jump Up to 200%
Updated
Updated · Network World · Aug 12

AI Memory Crunch Extends IT Lead Times to 18 Months as Prices Jump Up to 200%

3 articles · Updated · Network World · Aug 12

Summary

  • Lead times for servers, storage and networking gear have stretched from 30-45 days to 6, 12 or even 18 months as memory shortages choke enterprise IT projects.
  • Memory is the main bottleneck: Gartner says memory prices have risen 50%-200%, lifting PC prices 35%-45%, pushing some server prices up more than 125%, and driving network equipment prices over 20% higher in 2026.
  • Analysts and vendors blame hyperscalers absorbing memory capacity for AI, calling the disruption more durable than past shocks because the inference wave is only beginning and relief is unlikely before late 2027.
  • Enterprises are being urged to extend server life to 6-7 years, improve utilization, widen suppliers beyond preferred vendors, and use cloud or colocation capacity when on-prem hardware is unavailable.
  • Finance and procurement planning has become critical: experts recommend 12-24 month forecasts, frequent vendor coordination, and flexible monthly or quarterly purchases to avoid surprise budget overruns.

Insights

With hyperscalers hoarding hardware, how soon will the soaring costs of AI infrastructure trigger massive price hikes for everyday cloud services?
If the AI hype cools down, could today's severe hardware shortage suddenly collapse into the biggest tech surplus of the decade?