Buy servers now or cry later: DRAM price spike threatens infrastructure budgets

Buy servers now or cry later: DRAM price spike threatens infrastructure budgets

Summary

DRAM module prices jumped sharply between September and December 2025, with distribution data from analyst Context showing an average increase of around 63% for common capacities (16 GB, 32 GB, 64 GB and 128 GB) sold in Europe. OEMs such as HPE, Lenovo, Dell and Cisco have so far seen smaller increases (~28%) thanks to existing inventories and supply contracts, and complete server pricing only edged up 5–10% in December. However, vendors are burning through stock and are expected to buy replacement components at current inflated prices, which should push system pricing higher within the next three months. SSD and HDD prices have also risen roughly 30–40% since September, compounding the effect on total infrastructure costs.

Key Points

• Context’s distribution data: common DRAM modules up ~63% between Sep–Dec 2025 in Europe.
• OEMs have seen ~28% enterprise RAM price increases but have been cushioned by prior inventories and contracts.
• Server system pricing rose only 5–10% in December, but larger increases are expected as inventories run out.
• Storage components (SSDs/HDDs) have jumped ~30–40% since September, adding to infrastructure cost pressure.
• Enterprises may need to accelerate purchases, reforecast budgets or delay refreshes; cloud providers could pass costs on to customers.

Content Summary

The piece reports a fast-moving memory-price shock that began at the component level and is moving through the supply chain. Distribution metrics show the steepest increases for widely used DRAM module sizes; OEMs have so far managed a buffer thanks to inventory and contractual pricing, which has kept system-level price rises smaller for the moment. That buffer is temporary: as kit stocks are used up, vendors will purchase at current market prices and pass increases into rack-level servers, storage arrays and networking systems. Storage media has also seen notable inflation, meaning refresh projects and expansions face pressure on multiple fronts.

The article flags the timing risk for procurement teams: organisations planning infrastructure renewals should reassess budgets and schedules now, because buying later could mean paying materially more. It also notes implications for cloud operators — higher component costs for large datacentres could eventually flow through to service pricing for enterprise customers.

Context and Relevance

Why this matters: RAM and storage are baseline components of nearly every server purchase. Sudden supply-driven price swings translate quickly into higher capital and operating costs for IT departments and service providers. This story ties into broader industry trends: AI-driven demand, datacentre expansion and constrained semiconductor capacity have all pressured memory supply and pricing through 2025–26.

Who should care: CTOs, procurement leads, infrastructure and finance teams, MSPs and cloud customers — anyone planning hardware refreshes, capacity expansion or negotiating multi-year hosting contracts. Acting early could save meaningful budget and procurement headaches.

Why should I read this

Short and blunt: if you’re buying servers or planning infrastructure refreshes, this could cost you a lot more in a few months. Read now so you can decide whether to accelerate purchases, renegotiate timelines or adjust budgets. We’ve done the legwork — don’t be the last to find out your replacement kit just got pricier.

Source

Source: https://go.theregister.com/feed/www.theregister.com/2026/01/14/dram_infrastructure_costs/