Have we entered the age of circular energy?

For centuries, energy has been the epicenter of economies; from coal-powered factories to oil-fueled cars to today’s renewable sources. Recently, however, energy is discussed not in barrels of oil or the efficiency of solar panels, but in gigawatts of compute. The raw material fueling the AI landscape. 

 

Circular energy in this context has nothing to do with renewable sources, it is simply a term referring to the self reinforcing feedback loop driving billions of dollars into AI capital expenditure, and the two pioneers are chip manufacturing giants, Advance Micro Devices (AMD) and Nvidia. Brought together by the frontrunner of AI news - Open AI. 

 

To explain the sheer size of the deal between Open AI, AMD, and Nvidia, if the proposed data center plans run 24/7 the energy output produced would rival the gridlines of entire nations.

​

IMAGE IMAGE IMAGE

​

These massive deals aren’t just about chips, they signify a fundamental change in AI infrastructure economics, and energy sources globally. While the industrial revolution relied mainly on coal and oil, the new AI revolution is all about circular energy. 

 

Nvidia & The Everyday Consumer

NVIDIA and OpenAI recently signed a deal wherein OpenAI buys 10 gigawatts (GW) of computing power from NVIDIA, in exchange for NVIDIA investing 100 billion dollars into OpenAI. However, it is important to note that this 100 billion dollars is not upfront cash but rather provided as and when OpenAI buys NVIDIA chips to reach the 10 GW scale. Thus, introducing the circular component. Simply put, NVIDIA gives billions of dollars to OpenAI, only for OpenAI to use this investment (or at least a part of it) to buy NVIDIA’s own chips, which leads to an increase in AI capital expenditure and computing power, which poses the circular energy questions in the first place. 

 

But what are the implications of a massive deal like this? The impact goes far beyond Silicon Valley. 10 GW of computing power does not magically appear from thin air. According to CNN, Ashburn, Virginia, also known as “Data Center Alley”, has seen data centers built for AI compete directly with homes and factories for electricity; this has  put pressure on grids that were never designed for this level of demand. All of this has led to rising electricity bills for local households and businesses.

 

The NVIDIA–OpenAI deal is more than a commercial agreement; it illustrates how technology and energy economics are becoming interlinked. The commitment to 10 GW of compute will not only shape AI development and infrastructure but also directly relates to how power is allocated across grids. This is the essence of circular energy: capital flows into chips, chips drive demand for electricity, and that demand loops back into investment decisions. As data centers scale and utilities adjust, the challenge is for grid centers to meet rising energy needs without compromising reliability for households and industries. In this emerging AI landscape, development progress should not just be measured in model performance but in how efficiently we manage the energy cycle that sustains it.

 

AMD & Investors

While NVIDIA is by far the largest chipmaker, it is also imperative to consider one of its competitors, which is AMD. In October 2025, AMD and OpenAI settled on a multi‑year agreement for OpenAI to deploy up to 6 GW of AMD GPU capacity. In exchange, AMD issued OpenAI warrants to purchase up to 160 million shares (about 10% of the company) vesting as deployment and share‑price milestones are met. This deal was a hit across investors as AMD’s stock jumped over 20%. However, it is evident that yet again there is a circular structure as OpenAI gains GPU capacity, then effectively reinvests via stock-driven purchases which creates another loop of compute and capital.

 

The AMD–OpenAI deal signposts to other chipmakers and players in the AI market one clear message: to boost valuations, build, build, and build. Six gigawatts of GPU capacity is a deal that should be analyzed and critiqued from every angle possible before coming to a conclusion of its true value, but really it seemed like optimism was driving stock prices. An increase which will now just drive more capital expenditure. We have seemed to reach a constant cycle of loops. A concept quite difficult to wrap our heads around considering the fact that it underpins the daily necessity of energy. Circular energy clearly isn’t just a technical loop; it is also a financial one, where electricity demand and supply rise and fall with share prices.

 

What to make of it all? 

With these massive deals as well as countless other circular AI agreements happening everywhere, it is a challenge to comprehend what this all really means. For some clarity, these AI deals have turned energy and the infrastructure for it into a circular feedback loop. Coupled with this is the way these companies have been rewarded by the markets for the same deals. This works well for now: more energy for AI companies and more money for investors. Yet the moment these deals fail to meet expectations, as money flows away and projects get stranded, it is ultimately everyday consumers who suffer through rising electricity prices, a result far less impressive than these companies anticipate. So, to answer the question posed at the start: yes, we have entered the age of circular energy, and it is certainly not an age filled with security.