Bank of America: China's AI infrastructure capital expenditure is expected to reach US$1.5 trillion from 2026 to 2030, adding 35 to 45 GW of data center capacity

The Zhitong Finance App learned that Bank of America released a research report stating that it has launched the “Funding APAC's AI Revolution” (Funding APAC's AI Revolution) series of reports to explore investment in artificial intelligence (AI) infrastructure in the Asia-Pacific region and its impact on the credit market. The first report focuses on AI infrastructure construction in China. It is estimated that $1.5 trillion in capital expenditure (CAPEX) will be invested from 2026 to 2030 to increase 35 to 45 gigawatts (GW) of data center capacity and equipment upgrades. The Chinese market's capital expenditure per gigawatt is US$21 billion to US$29 billion per gigawatt, which is 40% to 45% lower than that of the US.

According to the Bank of America, based on 35 to 45 gigawatts of additional capacity and a median forecast of 21 billion to 29 billion US dollars per gigawatt of capital expenditure per gigawatt, the bank estimates growth capital expenditure of about 1 trillion US dollars. Additionally, the bank estimated maintenance capital expenditure of $500 billion. The estimate only covers IT equipment replacement expenses, is based on a 6-year renewal cycle, and excludes 30% of existing capacity that did not need to be updated due to vacancy.