How AI Could Make Your Next Phone or Laptop Much More Expensive
The surge of artificial intelligence is not only reshaping software, productivity and automation. It is also quietly rewriting the economics of hardware. From hyperscale data centres to the smartphone in your pocket, AI is emerging as a powerful force behind rising RAM prices worldwide, creating ripple effects across consumer electronics, capital markets and global supply chains.
At the heart of this shift lies an unprecedented investment cycle. According to estimates from Morgan Stanley, global investment in data centres could reach $3 trillion by 2030, while JPMorgan places the figure as high as $7 trillion. This race to build and expand AI-ready infrastructure is accelerating at a pace rarely seen before, and it is placing intense pressure on several critical bottlenecks: energy, infrastructure, chips, equipment and financing. Memory, particularly high-performance RAM, sits squarely in the middle of this storm.
The data centre boom behind the memory crunch
AI models are hungry for computing power, but they are equally dependent on fast, high-capacity memory. Training and running large language models, recommendation engines and generative systems requires massive amounts of RAM capable of handling parallel workloads with minimal latency.
This demand is being driven primarily by hyperscalers and cloud giants. From Google and Microsoft to OpenAI and Meta, companies are leasing and building enormous computing clusters optimized for AI workloads. These facilities require enterprise-grade memory, including high-bandwidth memory (HBM), which offers higher margins for manufacturers than consumer-grade RAM.
The scale of the buildout is staggering. In the United States alone, there are more than 5,400 data centres, according to Business Insider. That is more than Germany, the UK, China, France, Australia, the Netherlands, Russia, Japan and Brazil combined. States such as California, Texas, Illinois and Virginia have become focal points, supported by access to energy, digital connectivity and proximity to major technology firms.
According to Capital Group, this expansion represents a form of AI-driven reindustrialisation, extending well beyond the traditional tech sector into construction, industrial equipment, power generation and cooling systems. Memory suppliers are deeply embedded in this ecosystem, and demand from data centres is now reshaping their production priorities.
Why RAM prices are rising so fast
Over the past few months, RAM prices have more than doubled, and industry experts warn that the shortage could persist for several years. The issue is not simply a temporary imbalance. Memory makers such as Samsung and SK Hynix are struggling to increase output fast enough to meet data centre demand, particularly for advanced AI-focused memory.
Micron, the world’s third-largest memory manufacturer, recently announced the shutdown of its consumer-focused brand, citing a strategic shift toward serving AI companies that buy memory and storage in bulk. The logic is straightforward: enterprise-grade memory offers significantly higher profit margins, incentivising suppliers to prioritise these customers over consumer electronics manufacturers.
This prioritisation has tangible consequences. There have even been reports suggesting that Samsung’s memory division favoured external enterprise buyers over its own smartphone business, underscoring how intense the competition for memory resources has become.
Technician standing in server aisle while monitoring data systems on laptop. Surrounded by server racks, ensuring smooth operation of network systems.
Oracle as a warning signal
Oracle has emerged as something of a canary in the coal mine for the AI infrastructure boom. The company has faced investor scrutiny over its ability to finance and deliver large-scale data centre projects, including commitments linked to OpenAI expected to come online by 2027.
Oracle has denied rumours of delays caused by cost overruns or labour shortages, but the financial pressure is evident. Its committed order backlog has grown fivefold to more than $500 billion, forcing the company to issue nearly $30 billion in bonds, with further funding likely needed. Some of Oracle’s projects could require investments of up to $300 billion, highlighting the enormous capital intensity of AI infrastructure.
Financing the AI memory arms race
The magnitude of investment required is transforming how AI infrastructure is financed. According to Aberdeen, hyperscalers are increasingly turning to public bond markets as capital needs escalate. Until recently, these companies relied heavily on free cash flow and private funding.
That dynamic has changed rapidly. In just the past few months, Meta issued $30 billion in bonds, Alphabet $25 billion, Amazon $20 billion and Oracle $18 billion. Aberdeen expects this surge in issuance to continue, warning that repeated waves of supply could create periods of strain in public credit markets.
Building a data centre specifically optimized for AI can cost up to $50 billion, roughly three times the cost of a conventional facility. Add to this the rising energy requirements, and the financial burden grows even heavier. The International Energy Agency estimates that data centre electricity consumption will double by 2030, while AI-optimized servers could increase power usage fivefold. Bloomberg projects that by 2035, data centres could account for 4.4% of global energy demand.
Smartphones caught in the crossfire
While the epicentre of the memory shortage is in data centres, consumers are beginning to feel the impact. Smartphone manufacturers are facing higher component costs, and many are starting to pass these on to buyers.
Several brands have already raised prices or adjusted configurations. Devices that once launched with 8GB or 12GB of RAM are becoming more expensive, while higher-capacity variants are increasingly scarce. TrendForce reports that high-end smartphones may slow their transition to 16GB of RAM, potentially becoming rare by 2026. Some analysts suggest these versions could be quietly discontinued altogether.
Industry executives have been blunt about what lies ahead. Francis Wong, head of product marketing at Realme, has warned that prices are likely to keep rising from 2026 onwards, as reported by IndianExpress. Meanwhile, CMF by Nothing’s vice president of business, Himanshu Tandon, has predicted immediate price hikes across all segments, with flagship phones seeing increases of up to 10%.
As costs rise, manufacturers may respond by holding specifications steady across generations rather than pushing hardware upgrades. In the mid-range and budget segments, analysts expect RAM configurations to be capped lower, with a possible return of 4GB models driven less by preference and more by economic necessity.
A structural shift, not a short-term shock
What makes this cycle different from past memory shortages is its structural nature. The current pace of AI investment, according to Oxford Economics, mirrors the digital boom of the 1990s. AI is not a single product wave; it is an enabling technology spreading across industries, from cloud computing and finance to manufacturing and consumer devices.
As long as AI workloads continue to scale, competition between data centres and consumer electronics for memory will remain intense. With enterprise buyers willing to pay a premium, RAM is increasingly flowing toward the highest-margin applications.
The result is a new reality for the hardware market. RAM is no longer just a commodity shaped by consumer demand cycles. It has become a strategic resource, deeply tied to the economics of artificial intelligence and the infrastructure that supports it.
Frequently Asked Questions
Why is AI increasing ram prices?
AI workloads require large amounts of high-performance memory, driving demand from data centres and pushing prices higher.
Is the memory shortage temporary?
Most analysts believe the shortage could last several years due to sustained investment in AI infrastructure.
How does this affect smartphones?
Higher memory costs lead to increased phone prices and fewer high-RAM configurations, especially in mid-range devices.
Why are data centres prioritised over consumers?
Enterprise-grade memory offers higher margins, encouraging manufacturers to supply data centres first.
Will RAM prices keep rising?
While volatility remains, industry executives expect upward pressure on prices to continue as AI demand grows.
