How Are Data Centres Contributing To The Global Hard Drive Shortage?

A surge in demand for AI has shaken up the global storage market, forcing cloud service providers to turn to cheaper, faster options. With the supply of hard disk drives unable to meet demand, large cloud operators have begun using QLC (quad-level cell) SSDs for cold data storage.

QLC SSDs are cheaper than TLC (triple-level cell) models and perform well in systems that mostly read data instead of constantly rewriting it. This makes them a practical option for cloud services storing huge amounts of information that do not change often. The compromise is that QLC drives wear out faster, but their lower price and decent speed appeal to companies handling massive workloads.

According to DigiTimes, demand for these drives has outpaced production. NAND manufacturing lines producing QLC chips are fully booked until 2026. As cloud providers in the United States and China compete for the same limited supply, procurement prices are climbing. In response, major storage vendors such as Sandisk have raised SSD prices by as much as 50% after initially predicting smaller increases earlier in the quarter.

The global rush to build AI infrastructure has drained hardware supplies across the entire semiconductor sector . The race to reach artificial general intelligence has led companies to build data centres faster than manufacturers can produce the parts needed for them. Every tier of the supply chain, from GPUs and CPUs to DRAM and storage, is under strain.

Adata’s chairman told reporters that inventory buffers at semiconductor manufacturers have fallen from several months to only a few weeks. This means suppliers have almost no spare parts on hand to absorb new orders or delays. The pressure has left many smaller customers unable to source the components they need, while companies building AI systems are buying whatever stock remains, often at premium prices.

With production capacity tightening, the ripple effect reaches every corner of the market. Firms trying to purchase large-capacity HDDs are now waiting up to 24 months for delivery, according to DigiTimes. These drives, often used in enterprise-level nearline storage, have become almost impossible to get without a long delay. That has pushed hyperscalers to adopt QLC SSDs instead, worsening the shortage.

For chip and storage manufacturers, the shortage has created record profits. With AI companies ready to pay extra, suppliers are prioritising those large orders over smaller clients. Every major DRAM and NAND producer is now channelling resources toward AI-related contracts.

While the industry’s largest clients benefit from guaranteed access, the rest of the market faces delays and rising prices. Consumers are likely to see higher costs for everything from gaming PCs to budget SSDs through 2025. Even cheaper consumer drives that rely on QLC technology to stay affordable are expected to become more expensive in the coming months.

This situation has also exposed how unprepared the global storage market was for the current explosion of AI infrastructure. The 2 year backlog for enterprise HDDs shows that suppliers did not expect such a steep increase in demand. Expanding fabrication and packaging capacity takes years and large investment , meaning the shortage cannot be resolved quickly.

Related Articles News OpenAI Introduces New AI Agent That Conducts ‘Deep Research’ Zee February 3, 2025 News Elon Musk Wants To Buy OpenAI, Here’s How Altman Responded Zee February 11, 2025 News

Tourists Can Now Climb Mount Tai With AI Robotic Legs

Based on reporting by techround.co.uk.