Also an important part that keeps getting left out for some reason: they're doing so illegally. They are running unpermitted turbines. They claimed the turbines were off and journalists had to use thermal cameras mounted on drones to prove that they are indeed running. Then a non-profit had to sue them to get them to agree to turn off the turbines by 2027...
That's how bad our system is at giving any consequences to white collar crime.
and they (and many other AI data centers) run on some of the environmental/health hazard worst kinds of gas turbines available (i.e. cheap portable ones)
A well build gas power plane can be one of the cleanest fossile fuel power plants with neither huge issues of air nor sound pollution (compared to other fossile fuel, it still isn't perfect by far)
Same btw. for data center cooling, my office was for 3 years above a data center of a super computer in the mittle of a city and there where no sound issues at all (through yes it was a special case in many ways).
I mean yes, it costs more and many well established companies for high quality gas turbines are booked out.
Still using the kind of gas turbines which can bring back literal acid rain shouldn't be an option at all.
Lets also not speak about the environmental/health damage the production of the natural gas through fracking does.
Which bring us to a common nonsense argument, people going on and on about hoe economically expensive/harmful climate protection is, but totally missing that many oft he circumvented laws are not for climate protection but are for environmental protection for the sake of humans health.
To make thinks worse most of this data centers seems to be more about making money, then being in any way relevant for winning any AI race...
Problem is all the (relatively) efficient and (relatively) quiet turbines have long since been snapped up, so the AI sloppers have to go with inefficient, noisy ones for the foreseeable future. And given that no-one knows when the AI bubble will finally burst, no manufacturer wants to sink hundreds of millions into new plant to produce turbines in a year or two when the sloppers may be barely alive on taxpayer cash injections.
No, there are extreme consequences only when toes are stepped on. For example SBF... If it was run by Musk or Trump you can be sure there would be no sentence. I'm not arguing what SBF did isn't a crime. I'm saying if that deserves 20 years, then everything else should also be equally punished.
It’s an industrial supply bottleneck and rollout timeline incompatibility rather than a Dr Doom Evil Genius Complex that prevents everyone from deploying solar. The race dynamics make the discount rate on GPU deployments very high.
Solar is preferable when you have a ton of unproductive land and longer timelines to set it up - it’s certainly going to be cheaper than gas when everyone goes sets up NG plants as option 1 for new energy capacity.
The other confound is that China is the primary supplier of battery and solar tech, I’m sure that supply chain risk becomes part of the calculus when setting up so much infrastructure.
For scale, the current DC is about 8k acres, and the solar to run 7.6GW is almost 40k acres.
cheap, dirty, loud "tbh. shouldn't exist" kind fo gas generators are far more available and what is widely used
but also huge part of the solar supply bottleneck is artificial due to import regulations from china and both the US and EU having failed to secure meaningful domestic solar supply chains.
»It’s an industrial supply bottleneck and rollout timeline incompatibility rather than a Dr Doom Evil Genius Complex that prevents everyone from deploying solar.«
Idk. about any specific data center but many of this data centers aren't for AI training / making advancements on the AI race. They have sufficient resources for that already AFIK.
They are to run AI, so that the 100th app can force an AI agent down your throat weather your want or not /s. (Or in more optimistic words, so that the AI adoption can spread as fast as their is demand.)
The problem is that the grid can’t get built out fast enough. You’ll probably see most AI data centers switch to grid power in the coming years using mostly wind and solar because it’s simply cheaper as well
Remember when they said they couldnzt electrify the grid for ev's so they opted to just not do anything with ev's. Now datacenters are getting rushed in so fast and breaking the grids so hard they can illegally running generators
They could (and should be required to) build out their own solar/wind/battery infra as part of standing up a new datacenter. So much money is being wasted on poisoning our air and further enriching the richest companies in the history of the world.
You totally can. I've seen cases a rat gnawed a circuit breaker and blacked out a DC. Fires too. It's not an industrial manufacturing facility. Computers can just be shut off and on. There's also nothing needed to bootstrap the other parts like cooling etc.
A rat taking out a DC resulting in a black start event isn't the same thing as being able to turn a DC on and off at will. The thousands of computers in an DC running various bits of software don't just power on like your Windows computer and result in a working data center after a couple of minutes automatically.
Are the UPSes actual batteries or just backup generators? If they're generators then that defeats the point. If they're batteries then it's kinda just regular solar then?
But that's the point. It's not like a solar DC would suddenly lose power because of cloud cover. Large scale batteries can provide amble ride through inclement weather to allow a graceful shutdown, and weather forecasts at that scale are accurate enough to predictively model your energy availability.
AI training is unique because it's not interactive workload serving.
- Running billions of dollars worth of GPUs only 30-50% of the time insanely wasteful. Everyone else is running them at 99% duty cycle or close to it.
- Solar is big and generally has to be built far from where the data centres are. Transmission is already a bottleneck in many states. Adding tens of gigawatts requires non-trivial large scale infrastructure investment by… the utility companies, not SpaceX. Not to mention regulatory hurdles, etc.
- Battery production is already mostly maxed out, so obtaining tens of gigawatt hours of capacity on short notice is a decent percentage of global production.
Gas and other fossil fuel power plants can be deployed adjacent to the data centres for effectively zero transmission loss and no long powerlines that upset the NIMBY types.
They can work 24/7 and are largely available right now, but as others have pointed out the turbine blades are a limiting factor with these too.
> Solar is big and generally has to be built far from where the data centres are.
these data centers dont care that much about network throughput - you can sneakernet the training set into place and inference on the text isnt that taxing on the network
there's a reason the ai datacenters are being built wherever the power is cheapest and theres the lowest regulations. there's no real physical location requirement, so you could actually put them close to the solar power generation
Solar works as well as it does only because we’ve spread it out and connected it to a grid. When it is overcast in one spot it is sunny elsewhere.
That “elsewhere” has to be far enough away to get different weather.
This means a grid capable of transferring 100% of the needed power.
So even if you plopped a new DC in the middle of a field of panels you’d still need a ~1 GA grid connection… or a battery array many times the size as you’d otherwise need just to cover night time usage.
All those companies are sitting on literally hundreds of billions of dollars in cash. If they need this so badly, let them build the connection and other infrastructure.
They want to build the data centres right now, but the power companies and other energy suppliers are booked solid for 5+ years out.
Money or not, you can't just click your fingers and instantly get regulatory approval, or spin up a facility to build parts that only a handful of places have ever figured out how to mass-manufacture.
Gas turbine blades are especially difficult, but batteries and panels are not far behind. There are a surprisingly small number of shops capable of making power switching gear and transformers too.
You’re talking about one of the most hated person on hearth with one of the worst reputation. The literal Nazi salute guy. I don’t think he cares much about positive PR like that
Renewable energy only makes sense in a society that is agrarian. The second you have to make anything industrial you need baseload. That baseload can either be coal or nuclear.
In modern times you combine solar energy with wind power and batteries and you already have 90% of the load covered. Then combine that with cheap, easily adjustable gas power plants to complete the picture for cloudy, windless winter days.
There's really no need for coal or nuclear power plants anymore.
Building gas power plants that run only a few weeks per year is very expensive. Just compare prices in Germany (gas) with China (coal) or Finland (nuclear).
A more economical way would be to build nuclear power plants that power electrolyzers most of the time and provide power to the grid when necessary.
In the US, gas power plants make more sense due to ridiculously cheap gas, but even then the capacity factor of those power plants is much higher than 10%.
Comparing 24/7 AI training to operating hospitals 24/7 is a bit of a false equivalency, don't you think?
Besides, hydroelectric is commonly used as baseload where available. Additionally, there are big strides happening in power storage options to smooth the effective baseload of solar and wind.
I would argue that coal energy only makes sense in an anachronistic society doomed to fail.
You've been confidently wrong a couple times now - or maybe intentionally misrepresenting the facts?
In the US there are several states where hydro is the majority source of baseload electricity: Washington, Idaho, Vermont, Oregon, etc.
Outside of the US there are several countries that source more than 90% of their electricity from hydro - at higher percentages than Norway even: Paraguay, Albania, Democratic Republic of Congo, Nepal, Namibia, Zambia, Tajikistan.
You also don't have to build mountains to utilize hydro. There are areas that are utilizing pumped hydro to store the kinetic energy to smooth the baseline.
I've worked in energy and ran the numbers. Renewables only make sense for a pre industrial or deindustrializing society. When you need power to make shit they are useless. Which is why the industrial revolution moved from water to coal in the first place.
Yep and China is leading the world in recent and active construction of traditional hydro (dams) and pumped hydro storage to reduce their reliance on coal for baseload. I'm done with your bad-faith arguments.
Their 30 year plan is to move to nuclear for base load, and they are on track to manage it. The only people who think China is moving to a pre 2020 like western grid are people who don't know the first thing about China.
Or we can turn off the net negative data centers in order to keep our Hospitals running smoothly, rather than taking the entire grid down so some psychopaths in Silicon Valley can LARP at creating their silicon God.
* you can run data centres on grid electricity, which can come from mostly-renewable sources (in short, you need gas mainly as as an occasional backup).
* running things off-grid is dumb, and they’re only doing it because they’re desperate to get their data centres up and running yesterday so that they’re the ones who get to the singularity first (or some approximation thereof).
* the US electricity grid is f**d, mostly for the same reasons you can’t build any linear infrastructure in the United States - too many bodies hold veto powers and use any such infrastructure as an extortion opportunity.
* The spike in gas-powered generation for data centres is essentially a US phenomenon. Globally, the vast majority of new electricity supply is coming from renewables.
* in a sane world, Amazon and new large-scale data centre providers would face punitive carbon charges to incentivise them to switch to mostly renewables as soon as possible.
Gas is ridiculously cheap in the US while it's often the most expensive source of electricity in the rest of the world.
We're currently seeing a bifurcation where Asia bets on anything but oil and gas while the US goes in the opposite direction. Note that anything but oil and gas includes coal, so this isn't about the environment.
At least these sites (its more than Amazon) are being built near the energy source they will consume. Straight out of the ground into a turbine and then into the electron spinners.
Also there's not a lot of anything where these sites are located (Near El Paso). The 4-5 Road Runners who live around there will probably appreciate the change.
No one cares about West Texas, it's about the 33 million tons of carbon dioxide going into the atmosphere and warming the world.
Reminder that the USA are the country that historically burned the most fossil fuels, the world's largest extractor of oil, the only country trying to do a 180 on coal power.
Half of the world has experienced unprecedented heatwaves that will become the norm in 2050.
At the very least stop burning 3x more fossil fuels than the world average. Or stop your president from buying out renewable energy projects in order to prop up coal and gas. Or stop blocking public transportation developments. Why does it look like no one is trying?
The average grocery store food item travels more than 3k miles to reach its destination from the source. Let me know how you're going to make that work w/o diesel.
Electric trains can ship overland cheaper than diesel trucks. If we'd put half the money we put into free/subsidised highways into rail instead, the problem would already be solved.
The last mile could be fairly easily electrified at this point, battery trucks are quite sensible for short distances.
outside of the US most trains try to run electric when they can as it's just plain better (cheaper, more convenient too). Yes bunch of fright trains still run diesel, but that is mostly due to historic reasons and them often going to niche side tracks which have not been electrified. Fun fact a supper heavy mining fright train in Austrailia doesn't just run electric but basically 0-cost. It gets loaded in the mountains uses mostly gravity to run to the sea charging batteries while breaking and then when empty goes back up using the charges gained from breaking.
And electric trucks exist and work, too.
You don't even need latest battery tech as e.g. Swiss/Zürich hybrid bus/trams have been showing since some decades.
As another fun fact Beelitz-Heilstätten only allowed electric car around 1900!! (because it was healing lung issues and even way over 100 years ago it was well understood that air pollution from cars makes that worse).
Like, this was never about weather it's technical doable. But all about gigantic well established industries doing everything in their power to sabotage forced changes as such changes would make them obsolete. So a technical working solution isn't sufficient it must be noticeable cheaper or better so that it can disrupt the industry in a more "capitalist" way.
Even long-haul trucking can shrink rapidly between electric trucks (Australia is using those successfully - they’re cheaper and faster on hilly routes) on the shorter routes and electric trains (mature technology for a century) for many longer routes. Businesses would adjust but I think it’s really only aviation which is completely without a path until synthetic fuels arrive.
A better question is why a company is allowed to effectively lie using fraudulent false advertising in its marketing and brand image about being environmentally conscious and become the largest polluter in the country?
Is Amazon too broke to buy the necessary land to build out wind/solar/battery? I thought a company that name sponsored “Climate Pledge Arena” could afford to pay a little bit extra to care about the climate.
A wind/solar/battery solution is not even really much more expensive than gas power and it’s far less dependent on fluctuating fuel prices.
I'm absoltutely not defending the environmental decision, but these datacenter projects seem to be funded mostly by free cash flow (not borrowed money), from the hyperscaler profit margins.
I don’t think business really wants interest rates that high. Without conspiracy theories, a simpler explanation is that this is simply democracy in action. People don’t want the spending reigned in, and they don’t want their taxes raised.
If I understand correctly, for speed of deployment they're building a less-efficient simple-cycle plant (instead of the typical combined-cycle). Even though it's larger, the efficiency is worse than a typical natural gas plant.
Natural gas is basically a waste product that gets flared off in many oil-producing parts of Texas. If this private power plant can use that gas for something productive then I guess at least we're no worse off. But it's not clear that is what is happening.
Which is nonsense. There's plenty of pollution that's been diluted as much as it possibly can be and is still problematic. CO2 is an obvious example. Or consider leaded gasoline or mercury. You have to be careful not to eat too much fish or you'll have trouble with the mercury in it. That's been diluted by the entire ocean and that didn't solve it.
also, i don't necessarily mean to endorse this worldview in terms of its potential impact on policy, though there may be some truth to it, technically speaking.
The quip doesn't say "dilution helps." It says "the solution to pollution is dilution." Dilution solves the problem. And not only that, but it's the solution, not merely a solution.
This is just not true, and you seem to agree with that.
Yes, CO2 is problematic because of the sheer volume. The sheer volume means that dilution does not solve the problem.
Leaded gas and mercury are a problem because of bio-concentration. Which means that dilution does not solve the problem.
A bigger unit is generally more efficient, but this is a single plant with a bunch of units, and I don't know that the individual units are any bigger than they would be elsewhere.
https://terraformindustries.wordpress.com/2026/01/30/direct-...
Solar dc data centers are marginally cheaper. However right now the incentives are on buliding data centers (including power sources) the fastest. Despite this i would bet that you could build solar and batteries as fast as (or even faster) as building gas power plants, but the current governmend really likes to see fossils and setting the planet on fire.
Permitted maximum of 10 grams(2 nickels) of CO2 per hour for every person in the US. 33 million tons * 2000 ton/pound * 454 g/pound / 365 days in a year / 24 hours in a day / 342.6M people in us = 9.984 grams
It sounds... pretty good? It's not going to put pressure to the existing grid. Nor it will use freshwater. Near to the data centers themselves so the waste from power transmission is minimalized.
Except for that largest polluter in the country part. Our children and their children’s children and so on will curse our generation for what we did to them.
Yes, but those are a priority below massive solar and wind deployment because nuclear is another way to say emissions will rise for decades before the plants come online and start displacing fossil fuels. There are a few edge cases where nuclear might be necessary but there is an enormous amount of fossil fuel usage which could easily be replaced except for political capture by the oil and coal industries.
Making this argument is a strong indicator that you are extremely under-educated on this subject or motivated to make dishonest arguments.
There are a variety of commercial reactor design families that CANNOT melt down. Literally impossible for have a Chernobyl or Fukushima from the designs. These are not new. There is no reason anyone feeling like they know enough to contribute to this sort of discussion should be unaware. HTGRs, MSRs, SMRs, LMRs, SFRs...
"Sometimes nuclear plants melt down and cause widespread problems for very long timescales" is a nonsense argument to use against "We should build new reactors" when the new reactor designs we would be building can't melt down.
Nuclear is good for the last 10% (and very contextually - it maybe makes a lot of sense in Alaska but far less sense in Arizona, for example), but mostly we should be building more solar/wind/batteries, as they can be built both faster and increasingly cheaper over time (the cost of batteries has dropped ~97% in the last 20 years IIRC, solar and wind have similarly been dropping in price).
The main argument for nuclear besides "baseload" is that nuclear plants can potentially last for a century and will thus amortise its capex quite easily and afterwards become relatively cheap for most of its life. Except, given just how quickly solar is dropping right now, it's unreasonable to assume that 50+ years from now (so, in ~2076) already-built nuclear will be cheaper than contemporary solar tech (I.e. solar panels built in 2075). In fact, it's unreasonable to assume that nuclear plants will necessarily be longer-lasting than future solar panels, given current research into self-healing solar panels (which clearly isn't commercially ready, but very likely will be within 50 years). Even if they can only self-heal 3 times, that's 4x the lifetime for 120 years total, and much longer if the base lifetime of the panel is more than 30 years. And frankly nuclear will never beat solar panels on operational costs - nuclear has a 24/7 guard, a nuclear reactor and a turbine, whereas the closest thing PV has to moving parts is the de-dusting machines.
Similar logic for solar also applies to wind and batteries, of course - their price has been dropping like a rock for decades. And that compounds - if both solar and batteries get twice as cheap and their lifetime lasts twice as long, then the whole package becomes 4x cheaper. And the cheaper wind turbines get, the fewer batteries solar needs to plug the holes in intermittent generation.
A quick aside: a lot of focus on solar lifetimes assumes the minimum will be the average, but e.g. 1980s solar panels are still in operation with only 10-20% performance degradation IIRC so the popular assumption that solar panels last only 30 years seems a bit ridiculous.
China's nuclear stats should be kept in perspective by the way: IIRC ~90% of the generation they're building is non-nuclear renewables. They're also building more of basically everything than the rest of the world combined, so an absolute stat like "they're building 37 of X" isn't particularly useful except for conveying that China is building a lot of stuff.
To repeat, though: the US should still build nuclear plants (unless/until they get financially blitzed by renewables). My point here is to keep it in perspective.
And I don't mean to insinuate you said otherwise, I'm just kind of pre-emptively responding to a stereotype/common set of associated arguments, I guess.
So they can have cancer and not work... I guess that's better then working and having cancer. Or the people just starve instead so we don't have to worry about the pollution.
Carbon tax is the only way, you need to pay for picking up the garbage from the road or the air, the cost to clean op pollution and trash needs to be accounted for.
Currently the largest carbon capture and storage (CCS) project on the planet has failed to meet it's goals. Had it actually met the design goals it would be part of an LNG power source that manages to reclaim only 80% of the CO2 it releases.
That's not net zero, it falls short, and that's a good attempt but hardly viable.
The desired goal is to eliminate further emissions and also go further and claw back some of those gigatonnes of previously sequestered CO2 released by humans over the past century.
Carbon capture gets downvotes because it’s not working anywhere at scale. The fossil fuel industry loves it because it’s saying they get paid now and maybe someone else gets paid later, and their industry doesn’t have profits cut, but anyone planning to continue living on earth would prefer we stop polluting now because that is proven to actually work at scale.
What do you want? To keep chasing the perfect or to embrace digging ourselves out of this mess by whatever means? I think I know the answer and I'll bet we disagree.
But I too would prefer we pivoted to renewables and nuclear, increasingly renewables over nuclear with time, but I'd have to move to China for that (yet they ended up firing up the coal plants because of the Iran War). It's not happening here even when the free market indicates it's the best play because it undercuts oil and gas and they own the politicians as the masses mostly don't bother to vote. What is your plan that works in reality as opposed to the Aldous Huxley semi-dystopia the west has become? I know, I know, getting the job done by any means is wrong think, demanding perfection is right think.
I’m basing it on things like https://doi.org/10.1016/j.enconman.2026.121752 finding that it’s going to be hard both to scale up and get past 90% unless we have massive amounts of clean energy available. To me that means that we shouldn’t be approving any new fossil emissions and treating capture / sequestration as a nice bonus if we can make it work rather than something we can count on the make the numbers work.
The best thing I see is that economic pressure is mounting: even with Trump breaking so many laws on behalf of the fossil fuel industry, people want to deploy solar or wind power because that saves them money.
"The decarbonization of industrial hard-to-abate sectors through CCUS is critical, and that can be illustrated by looking at the target sources, such as the cement industry, accountable for 6–7% of global anthropogenic emissions [1]; or the steel manufacturing value chain, responsible for 7% of global energy-related emissions [2]. Given such dimensions, it can be concluded that carbon capture is critical to achieving CO2 restriction goals."
"In the cement industry it is worth highlighting that 60–70% of the total emissions of CO2 does not come from fuel combustion, but from the decomposition of the limestone into calcium oxide and CO2 "
So you can't get past 90% in the first place without carbon sequestration.
We have the same goal, but we see different paths to achieving it. I think if we do what it takes with carbon sequestration, the free market solves this for real as more and more profitable renewable power generation is stood up. But I do not believe that happens without the fossil fuel incumbents soft landing or finding bagholders to take the hit in the long run. And I won't count on Americans to demand better given all it took was the president to say he hates EVs and EV sales plummeted.
Personally, my house is more or less off the grid electrically already with a ton of solar panels and 75 kWH of battery backup. I voted with my wallet.
I'd love to hear how this will actually, really, pragmatically work.
Currently the largest Carbon capture and sequestration project on the planet is operated by Chevron as part of it's Gorgon Basin Natural Gas extraction project.
It's not working properly yet, when it is up and working as designed CO2 will be captured from the extracted natural gas and pumped back into the gas basin.
The total amount of CO2 captured and stashed will be a tiny tiny portion of the amount extracted and released.
That is very much not "Net Zero" no matter how much PR spin is applied.
Trees are wonderful for a number of reasons, and the need for them for local climate, which includes shading and their contribution to local humidity, is important as well.
They are not, though, a significant sequestration for CO₂.
A helpful way to think about this is to consider coal. To produce coal the energy input comes from time, pressure, etc over tens to thousands of millions of years.
The significance of trees is for the local climate, or global if you are considering one of handful of rainforests.
every time I see someone take seriously carbon credits/offsets I lose some hope. Possibly the most successful greenwashing campaign yet. Study after study have shown their uselessness. It doesn't even feel fair to call their problems as "loopholes" when, i.e., 90% of credits are completely worthless https://www.theguardian.com/environment/2023/jan/18/revealed...
The system isn't broken. It's working exactly as intended.
Yes, by burning fuels we release carbon from the thing being burned, it gets mixed with oxygen, and released into the atmosphere as co2.
As the suns energy warms the earth, the earth radiates heat back out and away from the surface and normally back out into space. co2 also acts like a blanket for that radiant heat, and traps it inside.
More co2 in the atmosphere traps more energy around the planet like an insulated wrap.
To reduce this, we have to extract the carbon out of the atmosphere and store it.
https://techcrunch.com/2026/08/07/spacexs-terafab-will-rely-...
That's how bad our system is at giving any consequences to white collar crime.
A well build gas power plane can be one of the cleanest fossile fuel power plants with neither huge issues of air nor sound pollution (compared to other fossile fuel, it still isn't perfect by far)
Same btw. for data center cooling, my office was for 3 years above a data center of a super computer in the mittle of a city and there where no sound issues at all (through yes it was a special case in many ways).
I mean yes, it costs more and many well established companies for high quality gas turbines are booked out.
Still using the kind of gas turbines which can bring back literal acid rain shouldn't be an option at all.
Lets also not speak about the environmental/health damage the production of the natural gas through fracking does.
Which bring us to a common nonsense argument, people going on and on about hoe economically expensive/harmful climate protection is, but totally missing that many oft he circumvented laws are not for climate protection but are for environmental protection for the sake of humans health.
To make thinks worse most of this data centers seems to be more about making money, then being in any way relevant for winning any AI race...
Especially if he buys from his own company.
Solar is preferable when you have a ton of unproductive land and longer timelines to set it up - it’s certainly going to be cheaper than gas when everyone goes sets up NG plants as option 1 for new energy capacity.
The other confound is that China is the primary supplier of battery and solar tech, I’m sure that supply chain risk becomes part of the calculus when setting up so much infrastructure.
For scale, the current DC is about 8k acres, and the solar to run 7.6GW is almost 40k acres.
cheap, dirty, loud "tbh. shouldn't exist" kind fo gas generators are far more available and what is widely used
but also huge part of the solar supply bottleneck is artificial due to import regulations from china and both the US and EU having failed to secure meaningful domestic solar supply chains.
No, it's just physics.
They are to run AI, so that the 100th app can force an AI agent down your throat weather your want or not /s. (Or in more optimistic words, so that the AI adoption can spread as fast as their is demand.)
Regarding EVs, the US started to fall behind in 2019, and the gap has only continued to widen, regardless of administration:
https://ourworldindata.org/grapher/electric-car-sales-share?...
Electric buses should be a no-brainer, yet New York comes last among US states in a recent market survey:
https://theicct.org/publication/zero-emission-bus-and-truck-...
Because you cannot turn a large datacenter with thousands of aggregates, power supplies and so on off and on like a toaster.
A 4 hour grid battery and a weather forecast would easily let a training run follow energy availability.
AI training is unique because it's not interactive workload serving.
i guess even at this stage “gas is easier” is still winning…
- Running billions of dollars worth of GPUs only 30-50% of the time insanely wasteful. Everyone else is running them at 99% duty cycle or close to it.
- Solar is big and generally has to be built far from where the data centres are. Transmission is already a bottleneck in many states. Adding tens of gigawatts requires non-trivial large scale infrastructure investment by… the utility companies, not SpaceX. Not to mention regulatory hurdles, etc.
- Battery production is already mostly maxed out, so obtaining tens of gigawatt hours of capacity on short notice is a decent percentage of global production.
Gas and other fossil fuel power plants can be deployed adjacent to the data centres for effectively zero transmission loss and no long powerlines that upset the NIMBY types.
They can work 24/7 and are largely available right now, but as others have pointed out the turbine blades are a limiting factor with these too.
these data centers dont care that much about network throughput - you can sneakernet the training set into place and inference on the text isnt that taxing on the network
there's a reason the ai datacenters are being built wherever the power is cheapest and theres the lowest regulations. there's no real physical location requirement, so you could actually put them close to the solar power generation
That “elsewhere” has to be far enough away to get different weather.
This means a grid capable of transferring 100% of the needed power.
So even if you plopped a new DC in the middle of a field of panels you’d still need a ~1 GA grid connection… or a battery array many times the size as you’d otherwise need just to cover night time usage.
They want to build the data centres right now, but the power companies and other energy suppliers are booked solid for 5+ years out.
Money or not, you can't just click your fingers and instantly get regulatory approval, or spin up a facility to build parts that only a handful of places have ever figured out how to mass-manufacture.
Gas turbine blades are especially difficult, but batteries and panels are not far behind. There are a surprisingly small number of shops capable of making power switching gear and transformers too.
And I want them to behave in a decent way, but universe does not ensure we get what we want.
They should be forced to build a grid in a way that does not speeds up global warming. What they want should be irrelevant.
Renewable energy only makes sense in a society that is agrarian. The second you have to make anything industrial you need baseload. That baseload can either be coal or nuclear.
There's really no need for coal or nuclear power plants anymore.
A more economical way would be to build nuclear power plants that power electrolyzers most of the time and provide power to the grid when necessary.
In the US, gas power plants make more sense due to ridiculously cheap gas, but even then the capacity factor of those power plants is much higher than 10%.
If you need to run an industrial process, not so much.
Besides, hydroelectric is commonly used as baseload where available. Additionally, there are big strides happening in power storage options to smooth the effective baseload of solar and wind.
I would argue that coal energy only makes sense in an anachronistic society doomed to fail.
In the US there are several states where hydro is the majority source of baseload electricity: Washington, Idaho, Vermont, Oregon, etc.
Outside of the US there are several countries that source more than 90% of their electricity from hydro - at higher percentages than Norway even: Paraguay, Albania, Democratic Republic of Congo, Nepal, Namibia, Zambia, Tajikistan.
You also don't have to build mountains to utilize hydro. There are areas that are utilizing pumped hydro to store the kinetic energy to smooth the baseline.
Quebec produces far more electricity via hydro than any single US state, some of which gets exported to the northeastern US.
Even China has seen a stark reduction in the percentage of electricity generated from coal plants over the past decade. Are they deindustrializing?
Everyone knows coal is a dead end. Maybe you should run the numbers again.
That's a funny statistic to focus on when the actual use of coal for electricity has increased by 50% over the same period.
It's almost like they are using it to cover baseload or something.
It shows that the percentage share of electricity generated from coal in China peaked at ~80% in 2007 and was at ~54% last year.
If unplugging a data center for five minutes can have a similar effect, then every AWS or cloudflare outage would have been a mass extinction event.
* you can run data centres on grid electricity, which can come from mostly-renewable sources (in short, you need gas mainly as as an occasional backup).
* running things off-grid is dumb, and they’re only doing it because they’re desperate to get their data centres up and running yesterday so that they’re the ones who get to the singularity first (or some approximation thereof).
* the US electricity grid is f**d, mostly for the same reasons you can’t build any linear infrastructure in the United States - too many bodies hold veto powers and use any such infrastructure as an extortion opportunity.
* The spike in gas-powered generation for data centres is essentially a US phenomenon. Globally, the vast majority of new electricity supply is coming from renewables.
* in a sane world, Amazon and new large-scale data centre providers would face punitive carbon charges to incentivise them to switch to mostly renewables as soon as possible.
We're currently seeing a bifurcation where Asia bets on anything but oil and gas while the US goes in the opposite direction. Note that anything but oil and gas includes coal, so this isn't about the environment.
The reason they're not doing that is because grid providers can't grow their supply fast enough to support new AI data centres.
256 comments, 8 hours ago
Also there's not a lot of anything where these sites are located (Near El Paso). The 4-5 Road Runners who live around there will probably appreciate the change.
West Texas is basically Mad Max 2.
Reminder that the USA are the country that historically burned the most fossil fuels, the world's largest extractor of oil, the only country trying to do a 180 on coal power.
Half of the world has experienced unprecedented heatwaves that will become the norm in 2050.
Stop burning fossil fuels.
The last mile could be fairly easily electrified at this point, battery trucks are quite sensible for short distances.
like
- electric trains
- electric cars
outside of the US most trains try to run electric when they can as it's just plain better (cheaper, more convenient too). Yes bunch of fright trains still run diesel, but that is mostly due to historic reasons and them often going to niche side tracks which have not been electrified. Fun fact a supper heavy mining fright train in Austrailia doesn't just run electric but basically 0-cost. It gets loaded in the mountains uses mostly gravity to run to the sea charging batteries while breaking and then when empty goes back up using the charges gained from breaking.
And electric trucks exist and work, too.
You don't even need latest battery tech as e.g. Swiss/Zürich hybrid bus/trams have been showing since some decades.
As another fun fact Beelitz-Heilstätten only allowed electric car around 1900!! (because it was healing lung issues and even way over 100 years ago it was well understood that air pollution from cars makes that worse).
Like, this was never about weather it's technical doable. But all about gigantic well established industries doing everything in their power to sabotage forced changes as such changes would make them obsolete. So a technical working solution isn't sufficient it must be noticeable cheaper or better so that it can disrupt the industry in a more "capitalist" way.
2. There are many alternatives to fossil diesel
3. The alternatives involved decentralised technology, so is actively opposed by entrenched interests who would burn the world for a dollar
For sure you can buy food grown within 500 miles, not delivered from other hemisphere.
https://www.thespruceeats.com/what-is-a-locavore-5118259
Hate urban modern life? Leave then. Lots of places to go live more simply.
It would be hard to decarbonize long haul trucking and aviation, which together account for less than ten percent of global oil use.
Almost everything else can be powered by other things.
Might have been true thirty or forty years ago, but even back then France had almost completely decarbonized the grid.
My guess would be a giant gas plant for a data center in Texas is part of the deal for letting them build it. Oil and gas owns that state.
Is Amazon too broke to buy the necessary land to build out wind/solar/battery? I thought a company that name sponsored “Climate Pledge Arena” could afford to pay a little bit extra to care about the climate.
A wind/solar/battery solution is not even really much more expensive than gas power and it’s far less dependent on fluctuating fuel prices.
It's more capital intensive, and interest rates are 8%, because our elected officials are completely incompetent.
Amazon can afford to power their data centers with clean energy.
It’s not called “Climate Pledge But Only If It’s Less Capital Intensive Arena”
I thought there was a gas turbine shortage?
Follow the money, and all that.
it's quite possible to have more product made, thus more pollution (or less materials consumed, thus less)
but concentration definitely will rise
also, i don't necessarily mean to endorse this worldview in terms of its potential impact on policy, though there may be some truth to it, technically speaking.
co2 is problematic because of concentration - the sheer volume.
leaded gas and mercury are a problem because of bio-concentration.
dilution still helps!
when people suffocate in the fire from CO2 - that's concentration
and dilution helps people to not literally die from that
global warming due to CO2 levels everywhere is entirely different problem. (and if we could dilute CO2 into the space, it would also be solved, lol)
This is just not true, and you seem to agree with that.
Yes, CO2 is problematic because of the sheer volume. The sheer volume means that dilution does not solve the problem.
Leaded gas and mercury are a problem because of bio-concentration. Which means that dilution does not solve the problem.
China is currently building 37 reactors simultaneously [1].
> [1] https://www.statista.com/statistics/513671/number-of-under-c...
There are a variety of commercial reactor design families that CANNOT melt down. Literally impossible for have a Chernobyl or Fukushima from the designs. These are not new. There is no reason anyone feeling like they know enough to contribute to this sort of discussion should be unaware. HTGRs, MSRs, SMRs, LMRs, SFRs...
"Sometimes nuclear plants melt down and cause widespread problems for very long timescales" is a nonsense argument to use against "We should build new reactors" when the new reactor designs we would be building can't melt down.
The main argument for nuclear besides "baseload" is that nuclear plants can potentially last for a century and will thus amortise its capex quite easily and afterwards become relatively cheap for most of its life. Except, given just how quickly solar is dropping right now, it's unreasonable to assume that 50+ years from now (so, in ~2076) already-built nuclear will be cheaper than contemporary solar tech (I.e. solar panels built in 2075). In fact, it's unreasonable to assume that nuclear plants will necessarily be longer-lasting than future solar panels, given current research into self-healing solar panels (which clearly isn't commercially ready, but very likely will be within 50 years). Even if they can only self-heal 3 times, that's 4x the lifetime for 120 years total, and much longer if the base lifetime of the panel is more than 30 years. And frankly nuclear will never beat solar panels on operational costs - nuclear has a 24/7 guard, a nuclear reactor and a turbine, whereas the closest thing PV has to moving parts is the de-dusting machines.
Similar logic for solar also applies to wind and batteries, of course - their price has been dropping like a rock for decades. And that compounds - if both solar and batteries get twice as cheap and their lifetime lasts twice as long, then the whole package becomes 4x cheaper. And the cheaper wind turbines get, the fewer batteries solar needs to plug the holes in intermittent generation.
A quick aside: a lot of focus on solar lifetimes assumes the minimum will be the average, but e.g. 1980s solar panels are still in operation with only 10-20% performance degradation IIRC so the popular assumption that solar panels last only 30 years seems a bit ridiculous.
China's nuclear stats should be kept in perspective by the way: IIRC ~90% of the generation they're building is non-nuclear renewables. They're also building more of basically everything than the rest of the world combined, so an absolute stat like "they're building 37 of X" isn't particularly useful except for conveying that China is building a lot of stuff.
To repeat, though: the US should still build nuclear plants (unless/until they get financially blitzed by renewables). My point here is to keep it in perspective.
And I don't mean to insinuate you said otherwise, I'm just kind of pre-emptively responding to a stereotype/common set of associated arguments, I guess.
Throughput for solar projects is very high and increasing, latency is unfortunately also very high.
It is too late now. There are no pare to optimal solutions left
https://www.aga.org/decarbonizing-through-carbon-capture/
https://www.energyindepth.org/study-finds-carbon-capture-and...
https://www.sciencedirect.com/science/article/pii/S277265682...
TIL mentioning a viable technology that would reduce emissions here but delay a forced transition to renewables is an instant downvote.
Currently the largest carbon capture and storage (CCS) project on the planet has failed to meet it's goals. Had it actually met the design goals it would be part of an LNG power source that manages to reclaim only 80% of the CO2 it releases.
That's not net zero, it falls short, and that's a good attempt but hardly viable.
The desired goal is to eliminate further emissions and also go further and claw back some of those gigatonnes of previously sequestered CO2 released by humans over the past century.
* A climate solution or distraction? - https://www.abc.net.au/news/2024-11-18/chevron-gorgon-fails-...
* ‘A shocking failure’: Chevron criticised for missing carbon capture target - https://www.theguardian.com/environment/2021/jul/20/a-shocki...
Also, I call BS on your claim it's not working anywhere at scale.
https://www.catf.us/resource/carbon-capture-storage-what-can...
What do you want? To keep chasing the perfect or to embrace digging ourselves out of this mess by whatever means? I think I know the answer and I'll bet we disagree.
But I too would prefer we pivoted to renewables and nuclear, increasingly renewables over nuclear with time, but I'd have to move to China for that (yet they ended up firing up the coal plants because of the Iran War). It's not happening here even when the free market indicates it's the best play because it undercuts oil and gas and they own the politicians as the masses mostly don't bother to vote. What is your plan that works in reality as opposed to the Aldous Huxley semi-dystopia the west has become? I know, I know, getting the job done by any means is wrong think, demanding perfection is right think.
The best thing I see is that economic pressure is mounting: even with Trump breaking so many laws on behalf of the fossil fuel industry, people want to deploy solar or wind power because that saves them money.
"The decarbonization of industrial hard-to-abate sectors through CCUS is critical, and that can be illustrated by looking at the target sources, such as the cement industry, accountable for 6–7% of global anthropogenic emissions [1]; or the steel manufacturing value chain, responsible for 7% of global energy-related emissions [2]. Given such dimensions, it can be concluded that carbon capture is critical to achieving CO2 restriction goals."
"In the cement industry it is worth highlighting that 60–70% of the total emissions of CO2 does not come from fuel combustion, but from the decomposition of the limestone into calcium oxide and CO2 "
So you can't get past 90% in the first place without carbon sequestration.
We have the same goal, but we see different paths to achieving it. I think if we do what it takes with carbon sequestration, the free market solves this for real as more and more profitable renewable power generation is stood up. But I do not believe that happens without the fossil fuel incumbents soft landing or finding bagholders to take the hit in the long run. And I won't count on Americans to demand better given all it took was the president to say he hates EVs and EV sales plummeted.
Personally, my house is more or less off the grid electrically already with a ton of solar panels and 75 kWH of battery backup. I voted with my wallet.
This is another level of insanity, right there.
Liars
Currently the largest Carbon capture and sequestration project on the planet is operated by Chevron as part of it's Gorgon Basin Natural Gas extraction project.
It's not working properly yet, when it is up and working as designed CO2 will be captured from the extracted natural gas and pumped back into the gas basin.
The total amount of CO2 captured and stashed will be a tiny tiny portion of the amount extracted and released.
That is very much not "Net Zero" no matter how much PR spin is applied.
They are not, though, a significant sequestration for CO₂.
A helpful way to think about this is to consider coal. To produce coal the energy input comes from time, pressure, etc over tens to thousands of millions of years.
The significance of trees is for the local climate, or global if you are considering one of handful of rainforests.
The system isn't broken. It's working exactly as intended.
It's just Amazon ~ Fuck Yeah We're Evil
As the suns energy warms the earth, the earth radiates heat back out and away from the surface and normally back out into space. co2 also acts like a blanket for that radiant heat, and traps it inside.
More co2 in the atmosphere traps more energy around the planet like an insulated wrap.
To reduce this, we have to extract the carbon out of the atmosphere and store it.