A growing shortage of memory chips is beginning to ripple across the economy, and the effects are reaching the infrastructure that powers the internet. As demand for artificial intelligence accelerates, manufacturers are shifting production toward high-performance chips used in data centers, tightening supply for the components that support everyday connectivity.
This shift is driving up costs for broadband equipment and raising concerns about long-term supply. As NCTA President and CEO Cory Gardner highlighted in a recent op-ed, the imbalance between AI demand and traditional infrastructure needs is becoming a critical challenge for U.S. technology leadership and network investment.
700–800%
Increase in DDR4 memory prices
Prices for DDR4 memory — a foundational component in routers, gateways and other broadband equipment — have surged 700–800% year-over-year. As manufacturers prioritize newer, higher-margin chips for AI applications, supply of legacy memory has tightened, driving sharp cost increases across industries.
70%
Projected share of global memory supply consumed by AI
Data centers already account for roughly half of global memory use, up from about one-third five years ago. That share is projected to rise further, with AI and data center demand expected to consume as much as 70% of global supply by 2026, tightening availability for the devices and infrastructure that power everyday connectivity.
20%
Share of router manufacturing costs tied to memory
Memory now accounts for more than 20% of the total manufacturing cost of a low-to-mid-range router, up from just 3% a year ago. This shift highlights how rising component costs are flowing through the supply chain, increasing the cost of equipment needed to build and maintain broadband networks.
Source: Counterpoint, Tom’s Hardware, PCWorld
