Artificial intelligence (AI) is being increasingly implemented across all industries, which is leading some to forecast semiconductor shortages. The demand for AI technology combined with global supply chain pressures like geopolitical tensions, natural disasters, and tariffs may challenge the industry’s ability to keep up.

The AI-Driven Demand Surge

Many businesses are rapidly adopting AI, with some of the top adopters being software development companies and customer support services.

There are two key areas in which demand is highest:

1. Data Centres and GPUs: AI models require graphics processing units (GPUs). By 2026 demand for GPUs is expected to rise 5-10% for data centres, while central processing unit (CPU) demand could increase by 1-5%.
2. Consumer Devices: AI-enabled personal devices, like smartphones and PCs, could further escalate demand. Microprocessor units (MPUs) in PCs are forecast to rise by 10-15%, and demand for smartphone system-on-a-chip (SoC) processors may grow by 15-20%. These spikes in both general and AI-specific components put additional strain on the semiconductor industry.

According to consulting firm Bain & Company, even a 20% demand increase could lead to a shortage market; current AI-related growth projections are likely to exceed this threshold. This could result in potential shortages in semiconductor components if the supply chain is unable to adapt.

Supply Chain Risks

While AI demand is the primary driver, other global issues could exacerbate any potential shortage:

• Geopolitical Tensions and Tariffs: These very topical issues are big influencers in current supply chain uncertainty.
• Natural Disasters: This unpredictable factor impacts the availability of raw materials like silicon and rare earth metals needed for semiconductor manufacturing. It can also impact fabs and disrupt production.
• Factory Construction: An increase in demand will require both larger manufacturing capacity and, especially in the case of data centres, larger facilities for the end product. However, any construction or expansion will take an extended period of time, making it impossible to instantly deal with heightened demand.
• Production Capacity: Data centre GPU demand could double by 2026, meaning suppliers would need to increase output by at least 30% to keep up. Insufficient production capacity could lead to a shortage in critical components.

The rapid growth of AI has affected two key markets, both relying on the semiconductor supply chain. Not only are data centres competing for GPUs to support AI models, but consumers upgrading to AI-capable devices are adding to the strain.

Some delays have already begun, with GPUs like NVIDIA’s Blackwell  out of stock until 2025. If this trend continues, consumers may experience further delays in obtaining AI-enabled devices, and companies could be forced to push back product releases.

Meeting AI Demand: Can the Industry Keep Up?

Although a shortage isn’t certain, geopolitical tensions and other factors could complicate the industry’s ability to maintain a steady supply. Businesses relying on consistent semiconductor availability may want to prepare for potential disruptions, such as buying stock in advance or scheduling deliveries.