AI & ComputeLOT 261119 · 3 min read

IEA Sees $4 Trillion Data Center Investment by 2030; IREN Jumps 7%

IREN rose 7% to $16.30 after the IEA forecast $4 trillion in added data center investment by 2030 and power demand doubling to 1,000 TWh, with grid and equipment shortages as the constraint.

By Grace Kim689 words

Key takeaways

  1. IEA's World Energy Outlook, published November 19, forecasts $4 trillion in additional data center investment by 2030
  2. Global data center electricity use to double to 1,000 TWh by 2030 under stated policies; AI drives 75% of the increase
  3. IREN rose 7% to $16.30; Q1 FY2027 AI Cloud revenue hit $200 million, up 182% sequentially, with a $20.3 billion backlog
  4. 93 purpose-built AI data centers tracked by the IEA hold 17 GW of inferred capacity, headed for about 50 GW when complete
  5. TeraWulf is up 11% YTD; IREN is down 57% and Applied Digital is down 37% year to date
IREN Gains 7% as IEA Sees $4 Trillion in Data Center Investment by 2030; Applied Digital and TeraWulf Rise 6%
IREN Gains 7% as IEA Sees $4 Trillion in Data Center Investment by 2030; Applied Digital and TeraWulf Rise 6% — Bru-nO / pixabay

The International Energy Agency expects $4 trillion in additional data center investment by 2030, and the forecast moved AI infrastructure stocks on November 19. IREN (NASDAQ: IREN) traded at $16.30, up 7% in morning trading. Applied Digital (NASDAQ: APLD) rose 6% to $15.35, and TeraWulf (NASDAQ: WULF) gained 6% to $12.72.

The IEA published the figure in its World Energy Outlook. The spending covers buildings, chips and electronics, and energy equipment including batteries, transformers, grids and gas turbines. For IREN, which holds more than 5 gigawatts of secured power, the report sharpens a question about whether a power-starved expansion favors operators that already control energy and sites.

The company owns land and grid-connected power across North America, Europe and Asia-Pacific. It is shifting from legacy crypto mining toward AI cloud services for customers such as Microsoft.

How much power will AI data centers need?

Spending is already running hot. The IEA put global data center investment at $565 billion in 2025. Electricity use at those facilities rose 17% that year.

Under its Stated Policies Scenario, the agency expects global data center electricity consumption to double to 1,000 terawatt-hours by 2030. AI-optimized facilities would drive 75% of that increase.

Purpose-built AI campuses carry much of the growth. Among the 93 purpose-built AI data centers the IEA tracks, inferred potential capacity reached 17 gigawatts as of September. Combined capacity is expected at roughly 50 gigawatts once the sites are completed.

Grids are where the forecast runs into trouble. Planned capacity in Northern Virginia, Texas, Denmark and Finland far exceeds what local grids can support, according to the agency. It also flagged shortages of advanced chip packaging, memory chips and power equipment as a speed limit on growth. The IEA stated plainly: "not all this prospective capacity will be built."

How did the three stocks diverge this year?

IREN's gain outpaced both peers, yet the stock remains down 57% year to date after a steep slide in recent sessions. On November 5, the company reported Q1 FY2027 AI Cloud revenue up 182% sequentially to $200 million. It posted a net loss of $263 million. Its revenue backlog grew to $20.3 billion.

Applied Digital trails on the year. The stock is down 37% year to date. The developer reported Q1 FY2027 results on October 7 showing 1.41 gigawatts of leased critical information technology load across five campuses.

TeraWulf is the group's outlier, up 11% year to date. Anthropic, its tenant at the Justified campus in Kentucky, priced its initial public offering above its range ahead of a Nasdaq debut. The pricing followed a losing week for the NASDAQ 100.

What separates the three business models?

IREN runs a vertically integrated AI cloud. It owns the data centers, the GPU compute and a software layer from its Mirantis acquisition, and rents out computing power directly.

Applied Digital operates more like a landlord. It designs and builds AI factory campuses for hyperscalers under long-term leases while also hosting crypto miners.

TeraWulf leans hardest on energy-market expertise. It leases critical IT capacity to AI tenants under long-duration, credit-backed leases as its legacy mining business contracts.

Applied Digital CEO Wes Cummins framed the scarcity argument all three share in the company's latest quarterly release: "Put simply, we view every new restriction elsewhere as making what we already own harder to replicate."

Can owned power carry IREN through the buildout?

IREN now faces the challenge of bringing contracted AI capacity online on schedule as grid and equipment bottlenecks tighten. The bullish case rests on the IEA's demand outlook and energized sites that newer entrants cannot easily replicate. Heavy capital spending and large debt loads leave all three companies exposed to construction delays.

The IEA's warning cuts both ways for IREN shareholders. Scarcity may raise the value of energized land, while shortages of power gear and memory chips may slow the projects meant to capture that demand. The timing of IREN's capacity deliveries could determine how much of the $4 trillion buildout the company ultimately captures.

247wallst.com (Source)

Filed under

Share this article:

More from Grace Kim

Grace Kim

Show full bio

Market editor covering people, launches and day-to-day practice in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,160 articles

Read also

« Previous articleNext article »