Santiago Pardo Sanchez

This week I did a re-read of Michael Liebreich’s magisterial summary of AI, data centers, and energy (“Generative AI – the Power and the Glory”) because I wanted to fine-tune the way I approached the subject at my day job.

Along the way, I kept notes and wanted to share them here. Having a basic understanding of AI, data centers, and the energy required is a requirement for informed citizenry because we are in the early stages of a transformational change last seen in the buildup to the industrial revolution.

I suggest you give the full article a read, but below are my (unrepresentative) excerpts and commentary.

——

This is also the year AI woke up to energy. The most powerful tech titans in the world have been humbled by the realization that their plans for world domination could be stymied by something as prosaic as electricity – and have embarked on a land grab for whatever sources of dispatchable power they can, triggering something of a gold rush.

This story is well reported, but strangely hasn’t sunk into the consciousness of the American people. Most in this country haven’t realized that the foundation for the future has been seized, claimed, and sold off. By the time public catches up, the second-order effects will be here and the winners will already be decided.

And then it dawned on everyone that the rate-limiting factor in the growth of AI was not going to be compute power, it was going to be electrical power.

Elon Musk, CEO of Tesla and owner of Twitter and X.ai, noted in March this year that “we have silicon shortage today, a transformer shortage in about a year, and electricity shortages in about two years”. Mark Zuckerberg, CEO of Meta, claimed his company would build more data centers if it could get more power.

Sam Altman, founder and CEO of OpenAI, summed up: “We still don’t appreciate the energy needs of this technology… there’s no way to get there without a breakthrough… we need fusion or we need radically cheaper solar plus storage, or something”. Those trying to find clean power for energy-intensive industries could be forgiven a little schadenfreude.

I’m a skeptic and think that the rate-limiting factor is labor (with government permitting as a potential third pending the actions of the unified Republican federal government). We don’t have the number of workers needed to build out the number of facilities that the hyperscalers have announced.

Alternatively, if you assume the $600 billion will be borne by just the world’s 100 million wealthiest and most intensive users of digital services, they or their employers would need to come up with a much larger $6,000 per year. Again, it’s not an impossible figure, but not one that can be found in next year’s budget. Meanwhile, planned capital expenditures by tech titans has continued to increase, so the required annual revenue per user keeps increasing.

The hyperscalers are all stuck in a prisoner’s dilemma. (I promise to go deeper into this at a future date.)

So, while it’s clear that AI is set to be truly transformative, it could still take a decade to justify current levels of investment. What that means is that there are two possible futures: One in which capital markets are happy to allow the hyperscalers to keep throwing money at AI in the expectation of future market leadership (and blowing out their balance sheets in the process), and one in which they are not.

The next recession will be a tech-led one and the bubble will pop once one of the hyperscalers misses a target.

However, the average data center has been getting bigger, and data centers are increasingly clustered onto campuses – two trends that put extreme pressure on the electrical grid. The Northern Virginia data center cluster, the largest in the world at around 2.5 gigawatts, soaks up around 20% of the region’s electrical power, growing at around 25% per year.

Why is the largest cluster of data centers in NOVA? Pure historical legacy and proof that American tech superiority pays off decades into the future. First, one of the first internet exchange points was funded by the Advanced Research Projects Agency (ARPA) in Arlington, VA. ARPA would create what would be called ARPANET, the precursor to the internet. This momentum would be followed by AOL and Equinox moving to the area and fully cemented in 1998 when MAE-East was put in NOVA. Think of MAE-East as a traffic intersection where highways (internet connections) come together, allowing data to travel faster and more directly between different providers. Keep in mind that this is a physical (to put it crudely) pipe network and that a majority of the world’s internet traffic flowed through it. This created the reinforcing loops that made NOVA the current and future leader of data centers in the world.

However, when it comes to “inference” – using the model to answer questions – results have to be delivered to the user without latency, and that means data centers in or near cities.

I wonder when we will get to a point where the investment in improved “internet pipes” will be a more economical move than trying to put the data center near the population center. Should we be buying land in 9th-best locations?

Chat GPT3 was trained using a cluster of 10,000 GPUs; Chat GPT4 required 25,000 GPUs; Chat GPT5 a rumored 50,000. Elon Musk’s x.AI has just built a 100,000-GPU data center, and there is already talk of the first million-GPU data center this side of 2030.

Elon’s work here was impressive from a logistical POV and snuck under the news radar because he was too busy helping out Trump. His x.AI went from zero to AI major in record time.

The most detailed demand analysis is by independent semiconductor supply chain specialists Semi Analysis. They have built a model based on analyst consensus for the number of GPUs in Nvidia’s sales pipeline, noting for instance that the three million AI accelerators expected to ship in 2024 alone would require up to 4.2GW of dispatchable power. Their base case has global data center demand tripling to around 1,500 terawatt-hours by 2030, growing from around 1.5% to 4.5% of total global power demand and requiring the building of 100GW of new dispatchable power supply.

Semi Analysis, however, goes one step further. They note that new AI data centers will not be evenly distributed around the world. They estimate that Europe today hosts only 4% of AI compute capacity; its commercial power prices are around double those of the US and, as Semi Analysis puts it, with some understatement, “adding a massive amount of power generation capacity to host the AI data center boom in Europe would be very challenging.”

Semi Analysis is the most detailed data center / semiconductor research shop in the game. Their analysis and deep dives are wonky to the n-th degree, but I highly recommend them.

Even if none of that happens, we could well see power demand limited by stunning improvements in energy efficiency. Nvidia claims to have delivered a 45,000 improvement in energy efficiency per token (a unit of data processed by AI models) over the past eight years. Their new Blackwell GPU is by some accounts 25 times more efficient than the previous generation.

Everyone always references the energy efficiency argument (A.K.A. Koomey’s Law [side note because I know the readership on this site is minimal: Koomey follows me on Bluesky and I think it was accidental because I am nobody]), but I think all analysis is underplaying it. Energy efficiency will lower the energy requirements more than we think!

For the US, I expect data-center capacity will somewhat more than double by 2030, adding around 30GW, and the rest of the world will add no more than 15GW.

Liebreich’s estimates seem okay to me, but I would make my estimate a bit lower than his. I fully agree with (and actively hope for) his ratio of the US vs World capacity. The incoming Trump administration will be extremely focused on ensuring that the ratio happens becomes a reality.

While AI will pose a huge challenge for the world’s power systems, it will also help mitigate energy demand in ways we can only begin to imagine.

AI is already driving a wave of innovation in the energy sector though, to be clear, not so much in the form of generative AI and Large Language Models. It is helping generators design, schedule maintenance, and optimize the output of wind and solar farms. It is being used to help grid operators squeeze more capacity out of existing transmission lines, to identify foliage encroachment on power lines, and to spot fires before they spread. AI is improving weather forecasting, which helps reduce peak heating and cooling loads. Leading utilities are using it to generate time-of-use price signals to send customers, helping to nudge demand to match periods of high renewable supply. AI-driven power trading is improving the utilization of grid-connected batteries and other resources. The list of potential use cases is endless.

In the broader economy, AI will help reduce pressure on energy demand. It will enable the optimization of freight and delivery networks, reducing congestion and fuel use. It will help optimize the utilization of infrastructure, reducing pressure on construction. Predictive maintenance will reduce costly and unnecessary repairs. Optimization of design and AI-driven new materials will reduce materials use across manufacturing. Every sector of the economy will be impacted. Eliminating road collisions, for instance, will reduce demand for health care and repair shops. Not to mention the reduction in power demand from replacing entire call centers with a few GPUs whirring away in a darkened data room.

“For the kingdom, the power, and the glory are [ours], now and forever. Amen.”

Could AI be so effective that it actually slows growth in demand for energy services? I doubt it. The more successful AI is, the more likely it is to trigger additional economic activity.

I guess? I’ll gladly temper my earlier comments with this caveat, but the premise can be solved with an energy abundance policy: with more renewables, we can have our cake and eat it.

When Elon Musk rushed to get x.AI’s Memphis Supercluster up and running in record time, he brought in 14 mobile natural gas-powered generators, each of them generating 2.5MW. It seems they do not require an air quality permit, as long as they do not remain in the same location for more than 364 days.

This is classic example of the genius of Elon.

Enhanced geothermal (based on fracking) and closed loop geothermal both look highly attractive because of their ability to deliver 24/7 clean power with the complexities of nuclear. Google and Meta have power purchase agreements with Fervo Energy and Sage Geosystems respectively. Other, more radical approaches, such as the millimeter wave drilling being proposed by MIT spinout Quaise, face daunting technological challenges and look decades away from commercialization, despite Quaise’s promise to provide first power by 2026.

2025 will be the year of geothermal. You heard it here third.

There is always good old hydro power. Grids with lots of it – like Scandinavia and Brazil – have always been good places to locate industries dependent on cheap 24/7 clean power, and they will be attractive to data-center operators. But it is hard and slow to build new hydro plants, and monopolizing the output of existing ones will be no more popular than in the case of nuclear power.

I’m secretly a hydro-skeptic for biodiversity reasons 🙈

Just as for other industries that risk imposing significant externalities on those around them, AI data-center owners will reduce cost and risk not by centralizing assets and building a wall around them, but by working with other stakeholders. Who knows, by leaning in, hyperscalers could even become popular contributors to the development of an affordable, resilient and green local grid.

This is the ultimate state of hyperscalers, but before we get there, many of them will make mistakes and we’ll see a new level of NIMBY battles.

When it comes to new technologies – be it SMRs, fusion, novel renewables or superconducting transmission lines – it is a blessing to have some cash-rich, technologically advanced, risk-tolerant players creating demand, which has for decades been missing in low-growth developed world power markets. Rather than focusing selfishly on funding them for their own use, however, the hyper scalers should help to underwrite their development and deployment in a grid-optimized way. Not only would this be more efficient for the overall economy, it could also help them secure their social license to operate, during what will be a period of extreme technological and social turbulence.

Firebrand British parliamentarian Tony Benn knew a thing or two about power. If we want to hold our leaders to account, he suggested, we should ask the following five questions:

“What power do you have? How did you get it? In whose interests do you exercise it? To whom are you accountable? And how can we get rid of you?”

When it comes to data centers, he certainly had a point.

Couldn’t have said it better.