IT News on October 2nd: On September 29th, according to Bloomberg, Bain & Company stated that to support the massive capital investment in current global data center construction, the global AI industry will need to generate $6 trillion in revenue annually by 2031.
The report released by this institution indicates that existing consumer-grade and enterprise-grade AI services can contribute at most $1.8 trillion, so there is still a need to find an additional $4.2 trillion in revenue. This additional revenue may come from new markets that are still in their infancy, including automated machines, robotics, as well as fields such as drug research and development, mental health, and energy production.
The main author of the report, David Crawford, Chairman of Bain's Global Technology, Media, and Telecommunications Practice, said: "The entire industry needs a wave of innovation that far exceeds that of mobile internet and cloud computing. AI Infrastructure construction is far ahead of demand, and to sustain these investments, the global GDP growth rate needs to increase by about 1% additionaly."
Bain's report suggests that to maintain the current development pace of AI, the industry still needs to overcome numerous obstacles.
Tech giants such as Microsoft, Google, Amazon, Meta, and Oracle are investing trillions of dollars in building data centers to meet the growing computing power demands of AI. The scale and construction costs of data centers double approximately every 12 to 16 months. The soaring prices of chips from manufacturers like NVIDIA and SK Hynix, as well as the increase in prices of network equipment and other components, are all contributing to this rise in costs.
At the same time, the debate over when AI service providers will be able to receive returns that match their substantial investments is growing day by day. Critics worry that the increasingly complex interests and supply relationships formed between technology device manufacturers and AI developers are driving up market expectations, and that these higher expectations in turn require more funds to support.
Bain believes that currently, the outside world is mainly concerned with to what extent AI can improve employee productivity, but the revenue from this alone is not sufficient to support the huge investments in the infrastructure of AI. The industry still needs to generate trillions of dollars in new revenue.
Bain predicts that by 2030, global data center investments will reach $5 trillion to $6.5 trillion, with an additional capacity of at least 150GW, further increasing the energy supply pressure on various countries.
By 2031, annual expenditures on infrastructure such as data center construction, computing power expansion, upgrades to AI accelerators, and storage chips could reach up to 1.5 trillion US dollars.
Data center developers are constrained by shortages of transformers, as well as water and electricity supplies, and at the same time face strong opposition from local residents. In the United States, during the quarter ending in June, data center projects that were thus halted or postponed amounted to a value of 68 billion US dollars.
