The AI boom and the crisis in the Middle East are driving up hardware prices.

Editor’s note: This article was updated on April 13, 2026, to reflect the current situation.

The AI supercycle is the latest major technological wave currently sweeping across nearly all industries, services, and devices. As AI rapidly evolves from specific applications to immersive and autonomous real-time systems, new demands are emerging that require massive amounts of hardware. Hyperscalers such as Microsoft, Amazon, Google, and Meta are spending billions on their AI infrastructures to train and operate models like GPT or advanced neural networks. Together, they plan to invest around $700 billion in AI infrastructure this year. In 2024, spending on new data centers totaled $250 to $290 billion, and in 2025, growth was at least 60 percent again. This massive investment boom in hardware and infrastructure will therefore continue this year as well[i]: semiconductors, manufacturing facilities, and power suppliers are at the center of it all.

Global supply chains under pressure, including in the Middle East

Investments are focused on energy-intensive systems that require enormous amounts of computing power. Manufacturers such as Nvidia, AMD, and Intel are prioritizing AI-specific components like GPUs and high-bandwidth memory (HBM) and scaling back production of conventional hardware. This is putting pressure on global supply chains.

The shift toward physical hardware makes AI as critical to the system as power grids or semiconductors. For companies, this development means higher costs for hardware procurement and longer delivery times, particularly for server- and storage-intensive systems.
Additionally, the conflict in the Middle East, which has been ongoing since March, is exacerbating the situation: Rising oil and energy prices, as well as disruptions in critical materials such as helium and bromine (important for chip production), are driving costs even higher and further extending delivery times. Qatar, which produces one-third of the world’s helium, had to temporarily halt production following attacks on gas facilities in March. Since Israel and Jordan together hold a massive share of the global bromine market, the tensions have also led to supply bottlenecks here[ii]. 

Price increases and bottlenecks

The AI boom is having a particularly severe impact on components such as memory (RAM/DRAM), processors (CPUs), storage systems (SSD/NAND), and network components (NICs). DRAM and NAND prices have already risen sharply; in the first quarter of 2026 alone, quarterly growth rates were 90–95 percent for DRAM and 55–60 percent for NAND. Analysts expect further increases in the second quarter of 58–63 percent for conventional DRAM and 70–75 percent for NAND flash. A 32-GB DDR4 kit, which was still available for $72 in 2025, now costs well over $260. SSD prices have more than tripled; a 30-TB enterprise SSD, which cost just over $3,000 in 2025, now runs about $11,000. The price increase for memory is unprecedented, Mike Howard, an analyst at research firm TechInsights, told the Wall Street Journal.

PCs, laptops and smartphones are becoming more expensive

Similar shortages are occurring with CPUs and network cards, as AI workloads require high bandwidth. TrendForce warns of “sharp price increases” due to AI demand. Short-notice orders are particularly affected, as hyperscalers take priority. Delivery times vary by manufacturer. The shortage could persist until 2027, as the industry has reduced its inventories to a buffer of just a few weeks. For end customers, this means more expensive PCs, laptops, and smartphones[iv]. Manufacturers such as Asus, Dell, HP Inc., and Lenovo have already announced price adjustments. Apple CEO Tim Cook also commented on this development during a conference call on the latest financial results, stating that Apple is facing bottlenecks in chip supply. Currently, rising energy prices resulting from the Middle East conflict are further driving up production costs.

Swiss perspective: opportunities in the crisis

In Switzerland, where innovation and precision are paramount, the AI supercycle presents both challenges and opportunities. Switzerland specializes in high-precision AI applications in key sectors such as medical technology, fintech, and environmental technology. A study by Google and digitalswitzerland predicts that AI could generate an annual value-added potential of up to 15 billion Swiss francs by 2034, primarily through efficiency gains in research and development[v]. Currently, however, the Swiss IT market is also feeling the impact of global shortages: companies such as banks and pharmaceutical firms are struggling with rising hardware costs and delivery delays. Swiss data center operators are investing in energy-efficient solutions, as the power demand from AI data centers is straining the grids[vi]. Furthermore, recent energy price hikes caused by the Middle East conflict are adding to the pressure.

Acting proactively for stable prices and deadlines

To mitigate the negative effects of the AI supercycle, we therefore recommend taking proactive steps (see box: What to do?). Plan your IT hardware procurement well in advance and implement your plans promptly in line with your current needs.

Contact UMB to plan your short- and medium-term procurement needs and work with us to define a strategy and solutions.