Home / Technology / AI Spending Creates Future Chip Demand Risk
AI Spending Creates Future Chip Demand Risk
11 Feb
Summary
- AI chip demand surge risks future data center idleness.
- Over $3 trillion projected for AI infrastructure in 5 years.
- High-bandwidth memory faces multi-year supply shortage.
China's leading chip manufacturer, SMIC, has cautioned that the current runaway spending on artificial intelligence chips is front-loading years of future demand. This aggressive investment in AI infrastructure, projected to surpass $3 trillion globally over the next five years, raises concerns that some data centers may become underutilized.
Companies are reportedly aiming to build a decade's worth of data center capacity within one to two years, without fully planning the operational needs for these facilities. This AI arms race is driving substantial capital expenditures, with tech giants like Alphabet, Amazon, Meta, and Microsoft collectively on track for $650 billion in 2026 alone.
The intense demand has also led to a critical shortage of high-bandwidth memory (HBM), a key component for advanced AI processing. SMIC notes that this supply crunch for HBM could continue for several years due to the time required to build and qualify new production capacity.
While China's AI developers, including Alibaba, Tencent, and ByteDance, are also investing heavily in domestic and foreign AI hardware, SMIC's production is limited by US export restrictions on cutting-edge manufacturing equipment, preventing it from producing the most advanced AI chips.
