Why the AI Boom Is Causing a Shortage of Mature‑Node Chips

The rapid growth of artificial intelligence is reshaping the semiconductor industry—but not only at the leading edge. While AI accelerators grab headlines, the boom is also creating pressure in the mature‑node semiconductor market, where most everyday chips are made. As global foundries redirect resources toward higher‑margin AI chips and memory, companies that rely on mature‑node production are increasingly turning to alternative suppliers such as China’s Semiconductor Manufacturing International Corp. (SMIC).

Why AI demand affects mature‑node chips

AI systems require a vast ecosystem of components beyond the main accelerator chips. These include power‑management ICs, controllers, sensors, analog chips, and other support components that are typically manufactured on mature nodes.

Despite the focus on advanced manufacturing, less than 20% of global semiconductor demand by product volume uses advanced nodes, while over 80% comes from mature‑node or specialty processes . That means even a shift in investment toward leading‑edge AI chips can have ripple effects across a much larger part of the industry.

As AI deployments scale, demand is rising not only for GPUs and specialized processors but also for the supporting chips that power and connect them. SMIC executives have said the expansion of AI applications is driving stronger demand for these supporting components, increasing orders and backlog for mature‑node manufacturing lines .

Foundries are prioritizing high‑margin AI components

Another reason mature‑node supply is tightening is simple economics. AI infrastructure—especially data‑center processors and memory—commands higher margins than many traditional semiconductor products.

As a result, foundries and memory manufacturers are increasingly allocating resources to AI‑related production. Some companies have redirected manufacturing capacity toward AI processors and high‑bandwidth memory (HBM), which are critical for training and running large AI models .

Industry analysts also report that manufacturers are shifting both 8‑inch and 12‑inch mature‑node capacity toward power‑related processes tied to AI infrastructure, which improves profitability per wafer . TrendForce estimates that the average utilization rate of 8‑inch foundry capacity among the world’s largest chipmakers could approach 90% in 2026 and remain above 80% into 2027, signaling a tight market for these older manufacturing lines .

A crowding‑out effect across the supply chain

Even when AI chips themselves use advanced nodes, their growth can crowd out other semiconductor production indirectly.

The semiconductor supply chain shares critical inputs—manufacturing equipment, materials, engineering talent, and investment capital. When AI demand surges, companies tend to prioritize projects tied to data‑center growth and high‑margin components. That can leave fewer resources available for chips used in consumer electronics, industrial equipment, and embedded systems .

The result is a structural imbalance: AI demand expands rapidly while production capacity for everyday chips grows more slowly.

Why Chinese foundries are receiving more orders

As overseas manufacturers shift toward AI‑focused production, some customers are looking elsewhere for mature‑node capacity. Chinese foundries—especially SMIC—have become an important alternative.

Industry reports indicate that the AI boom is pushing global customers to redirect orders to Chinese fabs as other foundries prioritize higher‑margin AI chips and memory . This dynamic has contributed to rising demand for SMIC’s mature‑node capacity, alongside higher prices and stronger utilization.

SMIC has reported that customer advance stockpiling and returning industrial orders are helping drive higher backlog and profitability as companies secure supply for future production . In practice, once buyers expect capacity to tighten, they often place larger or earlier orders to guarantee availability.

Localisation and geopolitical pressures

Supply‑chain localisation is reinforcing this shift. Governments and companies alike are trying to reduce dependence on a single region for critical semiconductor production.

Because most chips by volume are produced on mature nodes, domestic and regional foundries capable of manufacturing them have become strategically important. In China’s case, this has increased the relevance of local suppliers for industries ranging from consumer electronics to automotive components .

Industries most exposed to the shortage

The sectors most affected are those that depend heavily on mature‑node semiconductors. These include:

• Consumer electronics
• Internet of Things (IoT) devices
• Electric vehicles and automotive systems
• Industrial automation and robotics

Many products in these sectors rely on analog, power‑management, sensor, and microcontroller chips produced on older processes. When capacity tightens in those segments, manufacturers may face longer lead times or higher prices even if advanced AI chips remain available.

The key takeaway

The AI boom is transforming the semiconductor industry far beyond leading‑edge chips. By drawing investment and capacity toward high‑margin AI processors and memory, it is tightening supply for mature‑node manufacturing—where the majority of everyday electronics chips are produced.

As a result, companies across industries are diversifying their supply chains and increasingly placing orders with foundries that specialize in mature‑node production, including Chinese manufacturers such as SMIC. The severity of the shortage will likely vary by chip type and region, but the underlying dynamic—AI reshaping the entire semiconductor ecosystem—is already visible across the market.