4 moves supply chains must make as AI triggers a memory supply shock

AI has created a new supply chain constraint hiding in plain sight: memory.

The same DRAM and NAND components that power laptops, servers, storage arrays, smartphones, smart devices and networking equipment are now being pulled aggressively into AI data centers. As demand surges, memory prices have climbed sharply, putting pressure on enterprise technology budgets and procurement plans.

Gartner research finds that memory prices rose 50% to 200% in the first half of 2026, contributing to PC price increases of 35% to 45% and server price increases of more than 125% in some cases. Recent consumer electronics price moves show how quickly these market dynamics can ripple across the broader technology ecosystem.

The challenge extends beyond rising costs or limited availability. Organizations that move too slowly risk supply disruptions and delayed technology initiatives. Those that overcommit risk being locked into elevated costs as market conditions evolve.

The organizations that navigate this disruption most effectively will be those that balance supply assurance with commercial discipline, protecting critical technology roadmaps while avoiding unnecessary exposure to volatile pricing.

Here are four strategies to build resilience as memory shortages continue to reshape technology sourcing.

1. Forecast before the market moves

Traditional annual planning is too slow for the current market. Hardware quotes can expire in just a few days and configurations (and the inventory) can disappear before approvals are complete.

IT sourcing, procurement, infrastructure and enterprise architecture teams should build rolling 12- to 24-month hardware demand forecasts by category and business priority.

The goal is to create enough visibility to engage suppliers early, build an understanding of which SKUs may become difficult to obtain, and position buyers to capitalize on unexpected short-term availability when it arises. In a volatile market, forecasting is not just a planning discipline; it is a sourcing advantage.

2. Speed up buying decisions

When prices can shift between quote and shipment, delays in internal approvals become a financial risk.

Having preapproved contingency budgets, faster purchase-order authority and clearer escalation paths for constrained hardware categories can make a big difference. Procurement teams need the ability to act when inventory becomes available, rather than waiting through approval cycles designed for stable markets. Establish thresholds for acceptable price changes and require supplier notification of repricing or configuration changes.

In this environment, speed means staying ahead of disruptions, not panic buying. It’s about ensuring internal processes don’t amplify the effects of an already volatile market.

3. Build flexibility into specs

Fixed specifications are brittle in a constrained market.

If a program depends on one exact server, storage or endpoint configuration, it may stall when that configuration is unavailable, withdrawn or repriced. It’s important to focus less on a specific component or manufacturer for commoditized products and more on required outcomes: performance, capacity, resilience, security and lifecycle needs.

Technologists should also review their existing technology footprint and determine whether a refresh is the best option. For example, an organization facing constrained server supply may upgrade memory in existing equipment and postpone a noncritical refresh, preserving capacity while avoiding the cost and delays of scarce hardware.

While memory may also be constrained, upgrading existing equipment often requires sourcing only a single component rather than waiting for an entirely new server build that depends on multiple parts. The smaller physical footprint can also reduce transportation and supply chain complexity.

Organizations may also supplement capacity with older or refurbished equipment when appropriate.

4. Push back on pass-throughs

The memory shock will be priced into virtually everything.

As infrastructure costs rise, SaaS, IaaS and software providers may cite higher operating costs as justification for price increases. Some increases may be legitimate; others may reflect anticipated investments rather than actual customer consumption. Even legacy software providers may face increased hosting or data center expenses that are eventually passed on to customers.

Rather than accepting infrastructure-related price increases at face value, require vendors to demonstrate how higher costs are affecting the products and services being delivered, as it specifically relates to them.

Navigating the AI supply squeeze

AI is reshaping more than digital strategy. It’s driving up costs and creating supply constraints across both the physical and virtual digital supply chain.

The immediate priority is to make the organization more responsive than the market is volatile. This requires organizations to become more proactive in planning and more agile in how they evaluate sourcing and supplier decisions.

The organizations that emerge strongest from this disruption will be those that adapt quickly, limiting financial exposure while keeping strategic priorities on track.

Gartner analysts are providing further analysis on this topic at the Gartner Procurement Conference, taking place in San Diego, CA on Sept. 15-16.

About the author

James Smith is a Senior Director Analyst working with IT leaders around sourcing procurement and vendor management (SPVM) related issues in Gartner’s Supply Chain practice.

SC
MR