As top tech firms spend billions to develop and deploy generative artificial intelligence , the market for those products and services could reach almost $1 trillion in three years, according to a new report.

The total addressable market (TAM) for AI-related hardware and software is estimated to grow between 40% and 55% each year for at least the next three years, according to consulting firm Bain’s global technology report. That means the TAM — which refers to the potential overall revenue of a product or service if it has 100% share of the market — for AI could reach between $780 billion and $990 billion by 2027, according to Bain.

While there will be “fluctuations in supply and demand,” Bain sees a “long-term, durable trajectory” continuing for the foreseeable future.

“Companies are moving beyond the experimentation phase and are beginning to scale generative AI across the enterprise,” David Crawford, chairman of the global technology practice at Bain, said in a statement. “As they do, CIOs [chief innovation officers] will need to maintain production-grade AI solutions that will enable companies to adapt to a landscape that is quickly shifting. Essentially, they need to adopt an ‘AI everywhere’ approach.”

Bain identified “three centers of innovation” that will drive market growth, from cloud service providers leading the way with larger AI models and data centers, to enterprises and sovereign countries adopting AI technology.

As tech companies race to develop more powerful generative AI models, the report says “executives must now grapple with an additional layer of complexity and opportunity: the emergence of ‘sovereign’ AI blocs around the world.”

Due to COVID-19 pandemic disruptions and geopolitical tensions between the U.S. and China , companies have sought to make “more resilient” supply chains, leading to “new semiconductor hubs” in countries including India, Germany, and Japan, the report says. Now, the same shift is happening in other parts of the tech industry, according to the report, with governments around the world subsidizing sovereign AI by investing billions of dollars in “domestic computing infrastructure and AI models developed within their borders, trained on local data and languages.”

A growing AI market could lead to another period of chip shortages

Bain estimates the amount of AI processing could grow 25% to 35% annually over the next three years, which could lead to an increase in demand for large data centers in the next five to 10 years. Today’s data centers use about 50 to 200 megawatts of power and cost between $1 billion and $4 billion. In five years, the growth in AI workloads could spur data centers to grow to over a gigawatt, costing between $10 billion to $25 billion, the report says.

On top of data center growth will be the demand for graphics processing units (GPUs), which power AI models. The GPU demand “could increase total demand for certain upstream components by 30% or more by 2026,” Bain estimates.

Along with geopolitical tensions, Bain estimates “surging demand for AI computing power” will drive a chip shortage similar to the one caused by the COVID-19 pandemic.

For the latest news, Facebook , Twitter and Instagram .