Global technology companies are projected to spend more than $600 billion on artificial intelligence (AI) infrastructure in 2026, as the rapidly expanding AI Factory market moves into what analysts describe as an “industrialisation phase,” according to a new report by Omdia.
The market research firm said cumulative global investment in data centres could reach nearly $1.6 trillion by 2030, underscoring the scale of capital flowing into AI systems and computing infrastructure worldwide.
The report explained that AI factories are evolving beyond traditional data centres into large-scale industrial systems designed primarily to generate intelligence through advanced AI models and high-performance computing.
It described the shift as an irreversible transformation driven by rising demand for computing power, intensifying geopolitical competition, and the growing push for sovereign AI capabilities by governments and corporations.
According to Omdia, AI factories represent a new class of industrial infrastructure where the “output” is intelligence itself, with digital tokens and model performance increasingly defining productivity.
The firm noted that the sector is becoming increasingly capital-intensive, technically complex, and subject to tighter regulatory oversight as countries seek greater control over their digital ecosystems.
It identified five major trends shaping the AI infrastructure space in 2026, including a shift from raw GPU accumulation to performance efficiency, rising demand for sovereign and regional AI systems, wider adoption of liquid cooling technologies, expansion of specialised AI cloud providers, and stronger focus on long-term sustainability.
The report also highlighted a growing industry shift away from simply scaling up graphics processing units (GPUs), with firms now prioritising faster response times, cost efficiency, and operational optimisation.
Governments, it added, are increasingly investing in domestic AI infrastructure to reduce dependence on foreign technology providers and strengthen digital sovereignty.
Senior chief analyst for cloud and AI at Omdia, Raymond Zhan, said future competition in the sector would extend far beyond computing hardware and model development.
“The future competitive landscape will no longer be defined by model parameters or GPU counts alone,” Zhan said.
He explained that competition will increasingly depend on energy systems, cooling technologies, semiconductor supply chains, software ecosystems, regulatory compliance, and long-term capital resilience.
The report further revealed that the AI infrastructure ecosystem now includes hyperscale cloud operators, AI-focused cloud providers, private platform vendors, and regional operators serving industry-specific needs.
A survey of more than 200 organisations cited deployment delays, return-on-investment concerns, talent shortages, regulatory requirements, and engineering complexity as key challenges facing the industry.
Omdia projected that 2026 and 2027 will mark a critical growth phase for AI factories, with regional and sector-specific operators expected to drive the fastest expansion.
Industry analysts also expect the trend to significantly reshape future investments in cloud computing, telecommunications infrastructure, data centres, and AI-driven enterprise services globally, including in Africa.
