How many industries will Data Centres breathe life into?
TL;DR: At least 3 - Commercial Real Estate, Solar Power, Solar Module Manufacturing. And maybe Offshore Wind as well? There is, of course, nuclear!
(There’s some extrapolation, speculation, and cross-country pattern matching in this post. I’m not writing a policy brief, but considering what might happen in a world that no longer follows any kinds of trendlines. This is not me shilling data centres. And it’s most certainly not financial or investment advice.)
Commercial real estate globally suffered post the 2008 CDO crisis, picked up in the early 2010s, and then almost ground to a halt post the pandemic, when the largest buyers of office space - large corporates - put millions of square-feet premium space back on the market. And yet, real estate consultants and companies are the busiest they have been in decades.^1 There is a hunger to find just the right location for a new AI factory that nobody is quite able to sate.
Over the past decade, in India, massive (GW-scale) solar projects were set up through central govt (SECI) initiated tenders. But most of these started to see low bids in the last few years, due to poor off-take. At the end of 2024, SECI reported there was 9.2GW of tendered solar capacity yet to find power purchase agreements. In November 2025, many planned tenders, and many awarded tenders were cancelled by central govt agencies (SECI, SJVN, NTPC) due to lack of off-take. Several solar projects that were built in the past decade have experienced curtailment - unable to inject capacity into the grid due to grid capacity bottlenecks or having failed at finding buyers for full capacity. Moreover, this lack of off-take has been at a time when prices were the lowest they have ever been - as low as INR 2.42/kWh for pure solar and for solar + storage to INR 3.42/kWh.
But there’s been an interesting development this week in the US. Google, in order to power its upcoming data centers in Oklahoma (in Muskogee and Stillwater) will fully fund all infra needed to connect its data centers to the local utility, Oklahoma Gas & Electric (OG&E), grid. Google is also funding 2 new solar power projects to supply power to the data centers, with the agreement that extra capacity from the projects will be available for OG&E’s resource adequacy pool.
It seems to me that built, but yet untied solar capacity in India will soon be procured by data centre developers. (I gather this is cross-country extrapolation, but the demand in India - even if for inference and not training - will obviously outpace any other country simply due to the population numbers).
In addition to solar development, solar module manufacturing might also pick up again? China famously began to cancel expansion of module manufacturing capacity in 2024 (300GW of polysilicon and 200GW of wafer manufacturing projects cancelled as per IEA analysis), after module prices almost crashed in early 2024. Solar is certainly not the ideal solution for data centre energy (you want round the clock power, best served by nuclear power), but the cheapest and fastest to build currently, in line with the typical time it takes for a new data centre facility to be built.
The next industry I’m keeping an eye out for is offshore wind. Several times in the last 4 years, offshore wind tenders for areas off the coast of Scotland and eastern US saw underbidding simply because the cost of building offshore wind farms had risen incredibly. Everything from HVDC lines, to the specialised ships for installing offshore turbines, to the large turbines themselves had seen much cost inflation. Now that the end user (big tech and AI model companies), and not the utility, is directly making large power purchase agreements for energy capacity coming online in the next 3 years, I wonder if offshore wind will go from pilots to commercial scale before 2030?
^1 CBRE’s net revenue from facilities management went up by 24% in a single year in FY2024. And their overall revenue rose 13% Y-oY in FY25.
To track what’s happening data centres real-time see Telborg.
You can also read this post on the Telborg substack and subscribe to get posts in your inbox!
Full List of Sources for this post
Commercial Real Estate
India solar
China Solar Module Manufacturing
“Global solar manufacturing capacity is expected to reach over 1 100 GW by the end of 2024, more than double projected PV demand. This oversupply has caused module prices to more than halve since early 2023, leading to negative net margins for integrated solar PV manufacturers in 2024. The challenging market conditions have resulted in the cancellation of about 300 GW of polysilicon and 200 GW of wafer manufacturing capacity projects, valued at approximately USD 25 billion.” - IEA Renewables 2024 Report
Google–OG&E
Offshore wind
“At the same time, several offshore wind auctions were cancelled due to a lack of participation. Rising costs and overall uncertainty, paired with the absence of revenue stabilisation mechanisms such as contracts for difference, led to low interest from project developers. In June 2025, Estonia cancelled the Saare 7 seabed lease auction after both bidders failed to meet prequalification requirements. In August 2025, no bids were submitted in Germany’s auctions for two predeveloped sites totalling 2.5 GW of potential capacity. Similarly, no bidders participated in France’s second offshore wind auction in 2025. Denmark cancelled three auctions, while India abandoned two – one for 4 000 MW of seabed leases and the other for rights and support for a 500-MW offshore wind project in Gujarat – because of limited interest from developers.” - IEA Renewables 2025 Report