Skip to content
Sign in
Business · Global

BofA projects global data centre spending will reach $2.1 trillion by 2028

BofA Global Research estimates global data centre spending will reach $2.1 trillion by 2028, more than doubling from its projected $1 trillion in 2026. The firm says the market could grow to $2.93 trillion by 2030 as AI services drive demand and infrastructure development.

AI-assisted synthesis · Based on 4 independent reports from 5 outlets · Article text read from 1 of 4 reports · 1 article captured only in partHow it's made

View timeline · 6 entries
5 outlets2 regionsLatest reporting 2 min read
Coverage: Mostly CenterL 0 · C 4 · R 1
Updat3 graphic: Business
Updat3 graphic

What happened

Key facts

  • BofA Global Research estimates global data centre spending will reach $2.1 trillion by 2028 (per The Economic Times).
  • BofA Global Research estimates global data centre spending will reach $1 trillion in 2026 (per The Economic Times).
  • BofA says the data centre market, valued at $275 billion in 2022, could reach $2.93 trillion by 2030 (per The Economic Times).
  • Reuters reported that the estimates appeared in a BofA Global Research note on Friday (per The Economic Times).

The brief

BofA Global Research estimates that worldwide data centre spending will rise to $2.1 trillion by 2028, more than twice its projection of $1 trillion for 2026. The Economic Times reported the figures, which Reuters said appeared in a BofA research note on Friday. BofA links the forecast to growing demand for AI services and their development.

The firm also estimates that the broader data centre market, which it valued at $275 billion in 2022, could expand to $2.93 trillion by 2030. Its projections describe both spending and market value, distinct measures in the firm’s account. The brokerage estimates the market will grow at a 37% compound annual rate from 2025 through 2030.

It associates that forecast with AI-driven infrastructure expansion, while its stated explanation for the spending outlook is rising demand for and development of AI services. BofA puts construction costs at $49 million per megawatt and estimates that IT-related expenses make up 78% of overall spending.

The figures indicate the scale of investment in facilities and equipment embedded in the firm’s projections, but the report excerpt does not break down spending totals by individual companies or regions. For AI-focused data centres, which the account calls “neoclouds,” BofA estimates costs of about $57 million per megawatt.

It says these facilities need less space than conventional data centres, distinguishing them from the broader facility category in its construction-cost estimate. The Economic Times account attributes the forecasts to BofA Global Research and says Reuters reported the note’s estimates.

Newsmax, The Edge Malaysia, Reuters and CNA also ran headlines stating that data centre spending is expected to more than double to $2.1 trillion by 2028; their headlines do not add further details about the estimates. BofA’s projections extend from its 2026 spending estimate through its 2030 market-growth outlook, with AI services identified as the driver of the expansion.

Why it matters

  • BofA’s $2.1 trillion projection for 2028 signals the scale of spending the firm expects across data centres, compared with its $1 trillion estimate for 2026.
  • The firm estimates IT-related expenses account for 78% of total data centre spending, a substantial cost component in its forecast.
  • BofA estimates construction costs of $49 million per megawatt overall and about $57 million per megawatt for AI-focused facilities.
  • The forecast is tied to demand for and development of AI services, which BofA identifies as the driver of infrastructure expansion.

What to watch next

  • Whether BofA’s projected global spending reaches its $1 trillion estimate for 2026.
  • Whether spending approaches BofA’s $2.1 trillion projection by 2028.
  • Whether the data centre market tracks BofA’s projected 37% compound annual growth rate from 2025 to 2030.
  • Whether BofA’s projected market value of $2.93 trillion by 2030 is reached.

How it's being covered

Coverage analysisShow details
6 source articles · 5 unique outlets · 3 distinct reports
Outlet political mix: Mostly Center· 4 Center · 1 Right
Where the outlets are based: Asia 40% · US/Canada 20% · Other 40%
Outlet political mix
Mostly Center
Left
Center
Right
0 outlets
4 outlets
1 outlet
The marker shows coverage balance: each unique outlet counts once.
Geography
Where the outlets covering this story are based.
5 unique outletsDominant coverage: Asia
Coverage imbalance
  • Coverage is currently concentrated among Center-rated outlets (4 of 5 rated sources); comparatively few Left-rated outlets have published on the event.
How we measure coverage

One source is from a wire service (AP, Reuters, AFP or Bloomberg), whose reporting other outlets often republish.

How outlets' accounts compare

Updat3 does not yet have enough full reports to compare how outlets framed this story.

Timeline

Developments from this story's sources, dated background we compiled and fact-checked, and our earlier coverage. Each row says which.
  1. Equinix, founded in 1998, helped establish commercial data centres where networks and companies could interconnect.
    Background · from our historical context
  2. Amazon launched its cloud services, including EC2 and S3, in 2006, allowing businesses to rent computing and storage instead of building all their own facilities.
    Background · from our historical context
  3. AlexNet demonstrated the performance of a neural network trained on large datasets using GPUs.
    Background · compiled and fact-checked by Updat3
  4. Google researchers first submitted the paper “Attention Is All You Need,” which introduced the transformer architecture.
    Background · compiled and fact-checked by Updat3
  5. BofA valued the global data centre market at $275 billion.
    Background · compiled and fact-checked by Updat3
  6. On November 30, 2022, OpenAI released ChatGPT, bringing generative AI to a mass audience and helping drive demand for the computing capacity used to train and run AI models.
    Background · from our historical context

How we got here

On November 30, 2022, OpenAI released ChatGPT, bringing generative AI to a mass audience and helping drive demand for the computing capacity used to train and run AI models.

The technology’s modern architecture drew on the transformer design introduced by Google researchers in a 2017 paper; in 2012, the AlexNet system had already demonstrated the gains possible from training neural networks on large datasets with GPUs.

Original sources

1 of 6 articles · Global · Show all

Every story links to the reporting it was built from. How we choose sources

Spotted a factual error, a broken link or a misleading summary? Tell us.