Artificial Intelligence (AI) is driving one of the largest capital investment cycles the technology sector has ever seen. The key reason is simple: AI systems are incredibly compute-intensive. Training and running large-language models (LLMs) require specialised hardware – primarily semiconductors (but also others) on a massive scale.  

Capital expenditure by Big Tech firms reflects this urgency. The global AI data centre market is already worth over US $200bn and continues expanding rapidly, with hyperscalers investing billions annually to build out the necessary infrastructure.  

Meanwhile, overall data centre hardware investment is accelerating sharply, projected to exceed US $650bn this year.  

The central building block of AI infrastructure today is the graphics processing unit (GPU), a key type of semiconductor, which has evolved into the dominant engine for machine learning workloads (how AI is trained). Total semiconductor revenues are forecast to reach US $1.29 trillion in 2026, up 52.8% from last year.  

Another hardware component which has come into the spotlight is Memory, a core component in expanding an LLM’s ability to work with larger amounts of context (how far back it ‘remembers’ the conversation) and improving data transfer efficiency. The memory segment is also at the epicentre of the AI infrastructure buildout and total memory related revenues are projected to rise to US $594.7bn in 2026, driven by supply bottlenecks and surging demand.  

In the current phase of the AI cycle, hardware is capturing outsized economic value. This is characteristic of early-stage technology shifts – as AI software and platform monetisation matures, infrastructure providers and hardware manufacturers should continue to benefit from scarcity (limited supply), high switching costs (AI ecosystems tied to specific hardware platforms), and urgent demand (hyperscalers racing to build data centres and computing power capacity).  

While the funds and portfolios target growth themes like AI-Hardware, we aim to maintain a balance between growth and quality investing. We achieve this through prudent industry diversification and blending in defensive positioning through industries like healthcare and consumer staples, within our equity funds. 

What we’re reading: IDC - Semiconductor Market Forecast 2026: The AI Supercycle Arrives