Market research firm Gartner said in a report on the 14th that $3.5 trillion (about 4,713.8 trillion won) in AI infrastructure spending will face depreciation risk by 2030.
The firm's point is that capital spending to expand AI infrastructure is far outpacing the pace at which AI generates revenue, which will eventually cause a sharp drop in the value of that infrastructure.
Will Sommer, a senior analyst at Gartner, said in the report that hyperscalers are expected to spend about $1 trillion on AI infrastructure this year and will add $6 trillion more in investment from next year through 2030.
Revenue, however, is projected to fall well short of the scale of that capital spending. Sommer estimated that hyperscalers would need to generate $8.5 trillion in cumulative revenue by 2030 to cover the capital spending they have announced so far, along with financing costs and a minimum 12% return on invested capital. Actual revenue, he said, will reach at most $3.6 trillion. If capital spending continues at the current pace, the value of assets exposed to depreciation risk will exceed $4 trillion in 2032, he added. He projected that hyperscalers would be able to reach a 12% ROIC in 2032.
Sommer also predicted that more than 80% of neocloud providers will cease independent operations by 2030. Neoclouds are operators that provide cloud services built on graphics processing units and tailored to AI and high-performance computing. While hyperscalers have diversified sources of profitability, he explained, neoclouds will be hit harder by the imbalance between revenue and capital spending.
