The ecological and societal costs of data centers are hidden from the FAANG companies. It's very important to be well informed about it so society can regulate it. This podcast series, "Data Vampires" is really informative about the subject:
https://www.youtube.com/playlist?list=PLm-sqZXTqq9oIG_d0P7aT...
You can find it in your favorite podcast player. Everybody should listen to it.
Yes I'm sure "Data Vampires" is an unbiased evaluation of empirical evidence.
Datacenters are not appreciably different than other industrial operations in the scale of their water usage and I'm more curious about how this meme spread than about how evaporative cooling works.
If you don’t trust it, look up videos on YouTube of people whose water has changed color, or who have terrible noise pollution 24/7 now, or whose rural landscape got completely destroyed and paved over. There is real negative impact on people’s lives, ones who will never gain from tech companies getting rich.
Why would that matter? They aren't building massive multi billion dollar industrial facilities, they are building data centers. Or are you saying the impacts are justified if it is considered an industrial facility?
Because Louisiana is very rich with natural gas on cheap land, has low regulation, and governor after governor that is down for being a testing ground for new wacky republican policies.
Because it does? You either have to have a dispatchable energy source to turn on at night and on cloudy days (usually fossil fuels), or storage for 12+ hours. Actually more than 12 hours in case there's extended periods of cloud cover.
Zuck has aligned himself quickly (like the spineless ass kisser he is)
, ideologically and in action, to the Republican party. Of course Meta is gonna use fossil fuels and not woke solar.
While they are at it, I am pretty sure they will enable a couple of extra genocides like they did in Myanmar with the extra capacities from that data center.
Since it's over 3 hours total, here are the specific examples of "ecological costs" from this podcast:
* Data centers consume a lot of water. The example they start with is Google's data center in Dalles, Oregon, which used 355 million gallons of water in 2021. This amounted to 29% of all water consumed in the city (they did not bother to mention that the city's population is only 15K though).
* In 2023, hyperscale data centers used 66 billion liters of water in the U.S.: 3x the volume from ten years ago!
* They quote estimates that ChatGPT consumes 500 mL of water for every 10-50 user prompts (or 10-50 mL per prompt, which again sounds less dramatic).
* In Ireland, data centers collectively draw "over 20% of national electricity," which outstrips the total energy usage of all urban homes in the country.
* In Cerrillos, Chile, local residents blocked Google's plans to build a data center after discovering the scale of water use it would require (169 liters per second and its a drought-stricken area or something).
* The power consumption of data centers is enormous and places a very high load on energy grids worldwide. No specific numbers mentioned though.
My impression is that this is a clear example of a politically-biased podcast with an alarmist and accusatory tone, where none of the facts presented are particularly damning in the grand scheme of things.
In data centers with large water consumption, most of the water (90%) is used for evaporative cooling (letting hot water turn into vapor to carry away the heat), with the remaining 10% going to humidification systems (maintaining 40-60% humidity inside to prevent static electricity buildup, basically evaporation as well).
Let's take a moment to recognize what a dream "ecological cost" evaporating water is compared to old-time industries and the real environmental problems people have had to deal with. Old-timers in Cleveland can tell you how, until the 1970s (before the first serious ecological protection enforcement), the Cuyahoga River running through the city would CATCH FIRE and BURN Bible-style because of all the unprocessed, oil-based waste being dumped by plants and factories along its course. It is an unfortunate reality that many key industrial processes of our civilization dissolve dangerous and toxic compounds with water.
Also, the cost of evaporation cooling is not fundamental to data centers. You can change things around with some known engineering solutions and the costs for it would not be a deal breaker. For example, in Belgium, they built a two-loop water cooling system that can use industrial waste water (or even seawater in principle).
If you absolutely must, you can also build a fully closed-circuit liquid cooling system (think big fridge). The thing is that some water drawn from municipal system in a little city in the middle of nowhere isn't a problem.
As for high power consumption and "climate change impact," none of this is specific to data centers.
(And, this whole mindset about climate change in this podcast is just so 2010s. No, energy consumption is not inherently bad and sinful. No, the math of solving climate change with consuming less, putting on a sweater and saving does not work. A society that does not prioritize building for plentiful, cheap, (and yes,clean) energy is doomed to stagnate and wither economically. I see even most leftist people change their mind about this over the last few years, tired of never-ending green washing. If only political orthodoxies were able to change with the times...).
I was just looking at water usage in the Las Vegas area and found that one Google data center in Henderson consumes 320 million gallons a year. Just for comparison I looked at golf courses in the area and 1 golf course consumed 450 million gallons.
I see quite a few golf course near The Dalles, Oregon. (“The” is actually part of the town name).
The Dalles is just at the end of the wet side of the Columbia river gorge before it starts getting to a semi-arid region (starting about at Biggs Junction). Golf courses are more desirable toward the dryer side because it’s hard to get a game of golf in when it’s raining all the time. It is still a water limited place.
> Data centers consume a lot of water. The example they start with is Google's data center in Dalles, Oregon, which used 355 million gallons of water in 2021.
For comparison, one acre-foot of water is 325,850 gallons. Google’s data center used around 1090 acre-feet in 2021. One acre of alfalfa requires 4-6 acre-feet of water per harvest, so another way to look at it is Google’s data center used as much water as 218 acres of alfalfa. There are a million acres of alfalfa growing in California.
I'm not sure open loop cooling is really possible in those parts of the South. The idea that water consumption on the West Coast and South East is comparable is probably flawed.
I am not OP, but, sure, here are the specific facts on "ecological costs" from this podcast (with some help from Perplexity's amazing youtube summary feature):
• Data centers consume lots of water. The example they start with is that Google's data center in Dalles, Oregon used 355 million gallons of water in 2021, which was 29% of all water consumed in the city (its population is 15K though, but they neglect to mention that).
• By 2023, hyperscale data centers used 66 billion liters of water in the U.S.—triple their volume from less than a decade earlier.
• They quote estimates that ChattGPT consumes 500 mL of water for every 10-50 user prompts (or 10-50 mL per prompt).
• in Ireland datacenters in total draw "over 20% of national electricity", which outstrips the total energy usage of all urban homes in the country.
• In Cerrillos, Chile local residents blocked Google's plans to build a datacenter there after discovering the scale of water use the centers would require (169 liters per second iand its a drought-stricken area).
• Power consumption of data centers is enormous and puts a very high load on energy grids across the world.
My impression is that this is clear example of politically-biased podcast with alarmist and accusatory tone, where none of the facts presented are particularly damning in the grand scheme of things.
In data centers with large water consumption, the water mostly (90%) goes to evaporative cooling (they let hot water turn into vapor carrying away the heat) with the rest going to 10% humidification systems (getting to 40-60% humidity inside to prefect static electricity buildup).
Let's take a moment to recognize what a dream "ecological cost" it is - turning water into vapour - compared to the old timey industries and real environmental problems people have had to deal with. Old timers in Cleveland can tell you how until 1970s (before first serious ecological enforcement), Cuyahoga river running through the city would once in a while BURN WITH FIRE Bible-style from all kinds of unprocessed oil-based waste being dumped by plants and factories on its course. It's the unfortunate reality that many of our vital industrial processes that make our civilization possible rely on dissolving all kinds of most dangereous and toxic compounds in water.
Also, the cost of evaporation cooling is not something fundamental to data centers and is not something that cannot be altered with some known engineering solutions and manageble cost overheads if there is a need. For example, in Belgium they built a two-loop water cooling system that can use industry waste or even sea water. You can also get a fully closed-circuit zero running water cooling system (fridge-style) if you absolutely must.
As for high power consumption and "climate change impact", none of this is specific to data centers. We might like it or not but our society runs on energy and electric power. This whole mindset showcased in this podcast that all energy consumption is something bad and its all about reducing it is so 2010s. I think, its obvious by now that this the road to pure economical if not civilizational suicide. A society that does not prioritize building for plentiful, cheap and hopefully clean energy is doomed to wither and stagnate.
If solar is cheaper per watt than gas then they will build solar and use the gas plants to level the output. Assuming the grid capacity and/or nearby land is available.
Audio is just one form of media, though. I prefer reading as I can consume the information more quickly (and it tends to be denser and can be more precise and detailed), and am happy to pay for quality written content.
Paris Marx is writing a book on data centers, so you might get your chance yet! If you want to read about it now tho, he does cite all his sources and transcribes the episodes.
Good reporting would present the key parts of the article up top, but invite you to read the full story if interested.
The lead paragraph and the inverted pyramid are key concepts developed in the field of journalism.
As you might imagine, a two-hour-long podcast in a series named "Tech Won't Save Us" has time to explore many avenues of criticism.
Obviously, to summarise I have to remove the supporting examples, and the dozens of different people being interviewed. To be clear, the journalists aren't personally making all the criticisms, just interviewing other people, so if some of the following seems to contradict itself, that's why.
A fair chunk of the podcast involves explaining the context to a broad audience. You know, explaining what a data centre is, outlining the cloud market and its major players, etc.
The criticisms outlined in the podcast include:
* Data centres produce very few jobs for the communities they're located in.
* They are often built in struggling communities where 'enterprise zones' offer big tax breaks, hoping to attract employers.
* They consume quite a lot of power - not as much as, say, an aluminium smelter, but perhaps as much as 150,000 homes. Few cities have that much spare grid capacity, and some have warned about risks of rolling blackouts.
* 20 percent of Ireland's electricity is used for data centres (they're something of europe's data centre capital due to their attractive tax rates)
* Energy demand at data centres leads to greater emissions at power plants. Even if the data centre contracts to only buy renewable power, that might displace less-eco-friendly buyers of renewables onto non-renewable power sources. And a lot of things like 'carbon credits' are based on rather creative accounting.
* One high-profile data centre (in The Dalles, Oregon) is in a town suffering a drought, and consumes quite a lot of water, considering it's a drought area. Grass on the local golf course is completely dead.
* Land and tax breaks are often acquired through secretive shell companies that insist on secrecy agreements with desperate local governments; in one case the government didn't even know they were dealing with Google. This secrecy extends to agreements about things like water usage.
* As you can imagine, a local community suffering a drought sees the local data centre's water consumption being kept secret by elected officials, they assume the worst.
* Some data centre builders, like Elon Musk, have a history of making legally non-binding promises, then not bothering to keep them. And of running large gas generators without permits.
* The kind of distressed post-industrial communities that welcome data centres often have high levels of pollution and cancer, making those unpermitted generators particularly bad.
* Many of the hyperscalers are also big AI boosters, so it's not like the datacentre operators can disclaim responsibility for the power needs of AI.
* Many people have criticisms of AI, beyond energy consumption. Such as huge centralised LLMs transferring more control to huge tech firms; getting things wrong; AI friends being an alienating concept; having heavy-handed censorship; widespread use of bots on platforms like twitter and reddit; risks of job losses; being trained on pirated ebooks without authors' permission; being a really shitty therapist; producing mediocre art; producing porn depicting real people without their consent; producing creepy underage porn.
* Or AI might be a bubble that's about to burst, which would also be bad but for other reasons.
* Tech business leaders like Sam Altman are on record saying some pretty wacky things about AI power consumption, like that the high power demands of AI will force us to invent fusion power. A load of them also have weird, messianic ideas about "the singularity", or think we're all in a simulation already, or think living in a Matrix-style simulated 'metaverse' sounds like a great thing.
* Many of the highest-profile tech folks - the billionaires - have very right-wing politics. Such as opposing all regulation as a matter of principle, except on the occasions when it works for their benefit. Some people think expecting these folks to regulate themselves isn't the best idea.
Overall this is all stuff that followers of tech industry news will probably have heard before; the podcast just adds context, draws it together, and finds sources in the form of interviewees.
And yet computing power is what we need to advance technologicaly. Unstoppable force that indeed asks for solutions to new problems. Better acknowledge them then to fight them.
I feel like we need more social equality to advance technologically rather than gadgets. The idea that only the elites and rich can conjure technological wonders is just so demonstrably false and not needed in this moment of history.
Uplifting everyone ensures that we'll be that much more likely to find the next Mozart or Tesla or Torvalds or whoever, if we give them a chance.
But yes, better to acknowledge how capital can be better utilized. You can probably give away free school lunches for an entire generation of children with that $10 billion in Louisiana, or you can give it to Zuckerberg to get slightly richer.
Becomes abundantly clear which one is better for societal advancement.
Do you want one in your city? Taking water and power from your power infrastructure? I don’t, and thankfully our city government rejected an Amazon proposal to do that.
From one perspective, if they don't create many jobs and don't pay any taxes, I can see why communities don't welcome them.
From another perspective, compared to a coal mine or a paint factory or a steelworks or an airport or a landfill or an oil refinery, data centres are safe, low-pollution, low-noise, low-odour, and low-traffic. By the standards of industrial areas, they're great neighbours.
Great podcast, highly recommend it. Paris Marx is writing a book now on data centers. Between him and This Machine Kills it feels great to finally find some high quality tech journalism.
Access to very cheap power in the MISO region is likely one of the top driving factors for this location. It extends partially into Texas and I've found that my rates are sometimes as little as half of what ERCOT customers are paying.
The #1 thing that makes MISO so cheap is the fact that it has the heaviest coal generation mix (>40%) out of all US regional grid operators. Any talk about natural gas or renewables pales in comparison.
I would guess the 20 year 'tax break' (AKA the other taxpayers are footing the bill) is the real reason for the building. shell game
Meta built a data center in North Kansas City. I'm not sure details of their break (Mayor loves to hand out money), but power is likely cheaper, & def much greener (1/3 from wind farms in Western Kansas state last I checked).
"take advantage of a new Louisiana incentive program, established by Act 730, that offers qualifying projects a state and local sales and use tax rebate on the purchase or lease of data center equipment"
Curious as to how you reached this conclusion. No taxpayer funds are going towards construction or operation of the data center. The lack of tax revenue from Meta is nothing spent, and they're still going to be paying into the local economy. The energy infrastructure is going to be built by Entergy, who've projected it to cost customers ~$1 more or less per month.
As someone who lives here, this is one of the few times I agree with our government. We're one of the least competitive states in the country, our tech sector is almost non-existent. It's reasonable to offer what you can to attract business. I think Landry's LED efforts so far have been a respectable attempt at improving the state of things.
Because all of the plants are fully amortized. If they were brand new, the total cost would be excessive. Customers are mostly just paying for operations, fuel and maintenance. There is no capital recovery needed.
I'd be curious to test the GP's point. Since electricity maps doesn't have cost data for most US balancing authorities, you maybe could try figuring out power costs per balancing authority to end customers by using something like the https://www.eia.gov/electricity/data/eia861/ "Sales to Customers Customer Sited" data. Revenue over Megawatthours for Industrial service.
Louisiana is so historically corrupt : meta chose Louisiana because they can pretty much do what they want and the parishes (counties) like Richland where this is being built are so poor they couldnt care less about any environmental or quality of life setbacks. as a reminder , Louisiana is home to the cancer alley.https://en.m.wikipedia.org/wiki/Cancer_Alley
Reliability is king. Why not just buy 50 sq mi of ranch land in Arizona for a data center complex, install a solar and battery storage system and eliminate electricity concerns for the next 20 years? Using LFP or sodium-ion batteries for storage effectively eliminates fire risk.
Have there been any formal studies looking at OpEx ROI for offgrid carbon free generation + storage for data centers? Exiting the grid and vertically integrating on site generation eliminates a lot of risk when dealing with an external utility.
FAANG has the market cap to drive down costs for reliable carbon-free generation and storage, like Alphabet is doing[1].
Microsoft is demonstrating water-free cooling solutions[2]. As long as there's a fiber backbone nearby, FAANG can slash energy OpEx and not worry about the rest of the grid. Or natural gas prices.
While Meta has a non-binding promise to build more renewable energy, the Louisiana Legislature passed a new law that adds natural gas to the definition of green energy, allowing Zuckerberg and others to count Entergy’s gas turbines as “green.”
As much as I prefer burning gas over coal, conflating it with zero(-ish) emission energy sources like wind, solar, and nuclear is bad.
Due to all the methane leaks, gas isn't even as much cleaner than coal as it was purported to be... But hey monitoring programs got cut so I guess that solves the problem...
Anyone who has to live in a fairly closed system (i.e. this planet) in which fossil fuels are burned for power would be beyond a fool to not strongly prefer gas over coal seeing as their greenhouse emissions are close enough to be within arguing distance. It's all the other stuff coming out that's the problem with coal.
I think the point is: "you'd be a fool not to prefer gas, because while the greenhouse emissions are about the same, for everything else coal is much worse"
They said gas over coal. If you accept the claim that GHG emissions from gas and coal are roughly equal, their claim is the other pollutants from burning coal make gas far more preferable.
If their greenhouse emissions are even close only a moron would not pick gas over coal because the former's emissions lack all the other nasty byproducts that are present in the latter's emissions.
I agree methane leaks (and monitoring programs cuts) are a problem. But even with them, methane burns much more cleanly than coal. The former primarily emits CO2 and H2O, while the latter emits SO2, NOx, heavy metals and more.
These definitions always get muddled when flipping between CO2 emissions or pollution... coal is definitely worse from a pollution standpoint, is likely worse from a carbon standpoint, but much of the methane produced from natural gas production is just released into the atmosphere and has a dramatically higher warming effect compared to CO2 -- on the order of 80x more warming potential over 20 years and at least 20x over 100 years.
So only looking at the byproducts of methane combustion is also misleading since nat. gas plants largely aren't burning methane - and blanket statements for all natural gas are also misleading since e.g. the gas from Canada is extremely 'Sour' and releases a ton of sulfur compounds when burned, often with fewer scrubbers than coal plants.
Methane mostly disassembles into CO2 but it takes 12+ years. When thinking about global warming potential, everything is compared to CO2 which we’ve normalized as “1”. So something with a GWP of 2 is twice as bad as CO2 in equal volumes.
Methane will eventually break down into CO2, so if you look at the GWP for years 13-100, it’s 1. The weighted average for years 1-100 is over 20x, so it follows that if you look only at a shorter time frame, it would be dramatically higher and is indeed - somewhere north or 80 for a 20-year time frame.
As far as sour gas is concerned - not all natural gas formations are created equal. If you look at any serious pollution evaluation, they take into account which formation the gas was harvested from. Texas gas is pretty ‘sweet’ with low sulfur and acid content but much of the oil/gas in Western Canada or the Gulf is ‘sour’ and must be treated and refined prior to being sold as fuel. So it also follows here that flaring methane from sour fields is going to release a bunch of the souring compounds and have a much stronger environmental
impact as compared to sweet formations.
> Meta has a non-binding promise to build more renewable energy
Also the people working for that company. Unimaginable wealth, both at the corporate and personal level, everyone aware at this point that the climate is breaking down and yet, they just can't do the right thing because they are just too damn greedy.
I have to imagine it's just a complete lack of care and classifying it as "green" helps push through something that they're being lobbied to push. I can't imagine this is anything but nonsense.
Looks like Louisiana is all aboard the "internal colonialism" that seems to be all the rage at the state level lately. In this case, flouting national/international renewable energy policy so the good people of Louisiana can get the long term benefit of... Having to deal with the fallout of another datacentre project?
Come on Louisiana legislature, at least make them pay for resurfacing a highway or something.
Look, I love to be pedantic as much as the next person on this site, but let's not miss the forest for the trees. State level legislature relabeling fossil fuels so they count as "green" is not the path to a better future.
I'm not super-familiar with Louisiana, but my general impression is there's a lot of climate/weather events that are gonna impact power reliability. Hmmm.
If you've not paid attention to the recent hurricane damages to the US, it wasn't just coastal cities that were hammered. Lots of places "far enough away from the coast" saw lots of flooding. A hurricane doesn't just evaporate. The hurricane reverses the process back to Tropical Storm, Depression, etc while continuing to bring lots of rain minus all that wind
I'm assuming you're thinking about Helene. That was a very unusual situation that had a lot to do with local topography and geology on top of a very unusual weather situation (ground-saturating rains in the week prior to a hurricane).
This area hasn't been underwater in a meaningful way since the 1927 flood of the Mississippi.
Any situation that floods that area is going to be well beyond 100-year floods, likely beyond 1000-year, and (like the 1927 flood) predicated on catastrophic levee failures that result in the Mississippi being over 100 miles wide.
Is it possible? Of course. Is it likely? No. Would there be much bigger issues than a server farm being offline? Yes.
The worst-case scenario from a storm that hits locally (which, e.g., the 1927 flood was absolutely not a consequence of - it was water from upstream, which climate change will weaken because of decreased snowmelt) is like Harvey over Houston (similar terrain). But this is much farther inland and could not replenish itself from the Gulf.
> While Meta has a non-binding promise to build more renewable energy, the Louisiana Legislature passed a new law that adds natural gas to the definition of green energy, allowing Zuckerberg and others to count Entergy’s gas turbines as “green.”
This means it isn't securities fraud when Meta tries to meet "climate commitments" due to the greenwashing of fossil gas generation by the state of Louisiana. Louisiana is a low regulation jurisdiction that doesn't care if most of the state ends up a Superfund site, so it is ideal to colocate data centers that are going to burn up a bunch of fossil gas there over their lifetime (when they are unwelcome elsewhere).
While I like solar & wind, when oil is drilled, natural gas is often burned off at the drill site. So if it is going to be burned, might as well make electrify from it.
"World Bank is urging energy firms to gather the gas and sell it to businesses and consumers.... Companies can use the gas in mobile electricity generating stations, to power their oil drilling sites, or as a fuel in petrochemical plants."
A major Mississippi River flood could drown the site, but generally I’d think the risk level is fairly low.
It’s north of the Old River/Morganza flood control system. I think this would have to fail first in order for the Meta site to be in danger.
Also, if the New Madrid fault went off again, I think you’d kiss this site goodbye. But if that happened you’d have about a million higher priority concerns than a stupid data center.
I am amazed that it appears that Meta did not ask for tax breaks for a $10B project. Seems like they absolutely could have 'bid' this out among competing locales.
Thank you looks like they did get millions a year in no sales or use tax on data center equipment. Nothing compared to what NFL teams con out of cities for stadiums.
There is a big matrix of risk/reward for any DC location.
You bet Meta asked for incentives, but sometimes a guarantee of future power capacity, fast permitting, or ideal locations are worth more than the incentives the state could afford.
Sure, but you don't have to build your DC at the place that has the cheapest bid.
You just need to make them think you might not build somewhere else unless they sweeten the deal.
Hopefully this just means that governments have wisened up to the fact that a gazillion DCs are going to be built so if you pass on Meta you can just pickup Google's.
No. Meta is spending $10B to build its largest data center in rural Louisiana.
I guarantee you that a lot of that $10B will be spent out of state. This is yet another corporate handout with the thin veil of "technology investment" Louisiana loves.
Myth: It has computers, therefore we are investing in technology and hi-tech jobs.
Fact: This will be built by out of state contractors, staffed by mostly out of state workers, and far less than anyone expects or claims. And will essentially transfer local resources out of the state while making those resources more scarce for residents.
Louisiana let EA run their QA from here to severely underpay people and pay fewer taxes. They courted IBM to do the same with Salesforce jobs. And now Meta gets to exploit the state to enrich another out-of-state corporation.
I always told myself if I ever became a "tech billionaire" I'd buy out a random abandoned town somewhere, setup high speed internet, and turn a ghost town into a high tech town, cause why not? You could easily become mayor and approve some reasonable projects. Sell extremely affordable housing for the buck (close to actual cost).
I do often wonder if it might be worthwhile to shove a bunch of server farms into a few abandoned mines, if you setup the appropriate infrastructure in said mines to protect your data centers.
Disconnected from a major city, fiber internet, affordable, why would anyone want to move to Orlando it's all swamp lands? I remember that being Walt Disneys response when a reporter (who was correct) accused him of buying up all the land in Florida.
97% of the USA is already disconnected from a major city and affordable. And now with Starlink broadband internet is accessible as well. Yet there is no rush to migrate to rural areas - quite the opposite. People want to live in cities and suburbs.
The major tech companies are all scrambling to snap up cheap energy right now. The result is that we are dumping a whole lot of additional carbon in red states and adding a while lot of additional extremely expensive per MWh sources in blue states. In both cases, the winners will be tech company shareholders and the losers will be the people who actually live in these communities who will end up with dirtier, more expensive power.
The losers are going to be the energy companies who think they’re getting long-term energy sales but probably won’t be, since these techniques will get more efficient.
The techniques will get more efficient, but the quantity of training will increase monotonically. We aren't going to use less energy overall. The ratepayers are absolutely the ones who will lose out on this.
> Electricity demand in the U.S. held steady for 15 years but, last year, it increased by 3%— marking the fifth-highest rise this century. More jumps are projected for years to come.
Perhaps, my point was an electricity supplier that invested a ton of money in the early 2000s assuming that aggregate demand will keep growing forever would have been in for a rough time.
A variety of factors may or may not make a future where aggregate electricity demand would increases, or stagnates, or even declined.
Translation: Meta is wasting their money and destroying an already struggling community and the environment to be participating at something at which they are both hopelessly late to and hopelessly behind in.
Can't wait for their stock to crash and Zuck to be out.
I'll never understand why tech companies choose some of the locations for their data centers. Considering a big thing with data centers is "keeping stuff cool", you would think they would build them in the northern states, closer to Canada versus the hot sticky swamp.
> I'll never understand why tech companies choose some of the locations
That's because you've chosen not to read about it. Location is one of the most important things they think about for data centers and there are plenty of articles on the subject.
“We set out looking for a place where we could expand into gigawatts pretty quickly, and really get moving within that community on a large plot of land very quickly,” said Rachel Peterson, vice president of data centers for Meta. “We looked at finding very, very large contiguous plots of land that had access to the infrastructure that we need, the energy that we needed, and could move very, very quickly for us.”
To answer the question you're implying, surrounding temperature is pretty minor, the cooling required is orders of magnitude higher, so power access is more important; You'll frequently find them located near sources of energy.
Meta has defacto infinite money, they don't have to look for places where operation is cheap, but where they can be above the law as much as possible for doing whatever they want.
I'm pretty sure Meta's team has written about that at length. It's about many things, such as (power/transportation/internet/energy) infrastructure, political situation, available workforce, vicinity to population centers, property prices, and a whole lot more
Cheap gas, cheap land (it's on a big essentially empty plot that people have wanted to develop for a while, in a poor area with plenty of underutilized farmland), state and local governments that care more about this project than about environmental concerns.
Similar reason to why a lot of chemical manufacturing is in Louisiana.
Louisiana and New Orleans have been pushing to make the city a “tech hub” for the past 10 years (why would you build data centers in a flood-prone basin below sea level? I don’t know). I imagine most of it is striking a sweet heart deal with the municipalities that want the business.
There were several data centers and colos on Poydras St in downtown when Katrina hit. Famously SomethingAwful was being hosted out of one of them whose remaining on site employee live blogged the whole thing on LiveJournal.
I get where you're coming from but still I find funny in so many levels that the literal speed limit of the universe is too slow for our mundane (or even banal in FB case) needs. the universe isn't good enough to our need to move bullshit across the globe. surreal.
In the same vein it would be awesome if this _need for speed_ would materialize in infinite funding of neutrino based communication research.
> The project entails more than 2 gigawatts of computing capacity—Zuckerberg said it could eventually expand to 5 gigawatts—programmed to train open-source large language models.
Given that the human brain takes much longer to "train", I wonder how the energy efficiency pans out — comparing the two.
Biological systems are wildly energy efficient, that's kind of their whole thing. The average human will consume approximately 75kwh worth of calories in their lifetime. There are electric cars with bigger batteries.
[Edit] ok, yes, please. I get that i missed the k in kcal. The point stands. Biological training is massively more efficient, even when you forget to multiply by 1000
This is wrong by at least three orders of magnitude. Very roughly, a human requires 2000 kcal a day = 2 kWh a day so 75 kWh is enough to cover about a month, putting aside the upstream losses in the energy supply chain (which are far greater for humans).
In general, saying that biological systems are "wildly efficient" is... wildly wrong. Some biological processes are optimized by evolution... most are not. There are no bicycles in nature.
> The average human will consume approximately 75kwh worth of calories in their lifetime. There are electric cars with bigger batteries.
Doesn't pass the smell test. I think I could push an electric car at least a mile a day if that's what I spent most my extra calories on. If I did that I'd surpass its range in well under 2 years, much less than my lifetime.
We're efficient once we have the energy, sure. How much energy does it take to go from raw sunlight to a calorie your body is actually able to use, and finally to your dinner table?
All of our food was alive before we ate it. All calories used by living things are efficient.
Life is an end unto itself. It does not need to justify its existence by the moral code of technocrat materialism. The fact that this discussion is being had on this board in good faith is morally condemning of our worldview.
Since the original point of this chain was a comparison between the energy efficiency of biological vs machine learning, then we need to be trying to understand if the machine is more efficient than the human. You don't need to make some moral or philosophical argument about existential justification to accept that taking a more efficient approach is better, in that it generally enables more life for the same energy.
If the true, total cost of a machine to perform some task is less than a person to do the same task, then the machine should do it and the person should move to do what the machine cannot. This means more energy is available for everything else, living included.
The ecological and societal costs of data centers are hidden from the FAANG companies. It's very important to be well informed about it so society can regulate it. This podcast series, "Data Vampires" is really informative about the subject: https://www.youtube.com/playlist?list=PLm-sqZXTqq9oIG_d0P7aT...
You can find it in your favorite podcast player. Everybody should listen to it.
Yes I'm sure "Data Vampires" is an unbiased evaluation of empirical evidence.
Datacenters are not appreciably different than other industrial operations in the scale of their water usage and I'm more curious about how this meme spread than about how evaporative cooling works.
If you don’t trust it, look up videos on YouTube of people whose water has changed color, or who have terrible noise pollution 24/7 now, or whose rural landscape got completely destroyed and paved over. There is real negative impact on people’s lives, ones who will never gain from tech companies getting rich.
Again, is there anything distinctive about data centers, or this industrial facilities in general?
Why would that matter? They aren't building massive multi billion dollar industrial facilities, they are building data centers. Or are you saying the impacts are justified if it is considered an industrial facility?
I am not sure why the question is relevant to my comment. If it weren’t distinctive, what does that change about it being harmful and unfair?
Why Louisiana? It's pretty hot most of the year. Why not northern states or even Canada?
And burning fossil fuels is pretty shady considering how cheap solar has gotten.
Because Louisiana is very rich with natural gas on cheap land, has low regulation, and governor after governor that is down for being a testing ground for new wacky republican policies.
They also placed it near Grand Gulf, and there is a lot of water in that area.
People still want their software to work at night. And last I checked, Louisiana had access to lots of water.
Wait, are you making a “with solar electricity will go away at night” argument?
Because it does? You either have to have a dispatchable energy source to turn on at night and on cloudy days (usually fossil fuels), or storage for 12+ hours. Actually more than 12 hours in case there's extended periods of cloud cover.
Zuck has aligned himself quickly (like the spineless ass kisser he is) , ideologically and in action, to the Republican party. Of course Meta is gonna use fossil fuels and not woke solar.
While they are at it, I am pretty sure they will enable a couple of extra genocides like they did in Myanmar with the extra capacities from that data center.
[dead]
Since it's over 3 hours total, here are the specific examples of "ecological costs" from this podcast:
* Data centers consume a lot of water. The example they start with is Google's data center in Dalles, Oregon, which used 355 million gallons of water in 2021. This amounted to 29% of all water consumed in the city (they did not bother to mention that the city's population is only 15K though).
* In 2023, hyperscale data centers used 66 billion liters of water in the U.S.: 3x the volume from ten years ago!
* They quote estimates that ChatGPT consumes 500 mL of water for every 10-50 user prompts (or 10-50 mL per prompt, which again sounds less dramatic).
* In Ireland, data centers collectively draw "over 20% of national electricity," which outstrips the total energy usage of all urban homes in the country.
* In Cerrillos, Chile, local residents blocked Google's plans to build a data center after discovering the scale of water use it would require (169 liters per second and its a drought-stricken area or something).
* The power consumption of data centers is enormous and places a very high load on energy grids worldwide. No specific numbers mentioned though.
My impression is that this is a clear example of a politically-biased podcast with an alarmist and accusatory tone, where none of the facts presented are particularly damning in the grand scheme of things.
In data centers with large water consumption, most of the water (90%) is used for evaporative cooling (letting hot water turn into vapor to carry away the heat), with the remaining 10% going to humidification systems (maintaining 40-60% humidity inside to prevent static electricity buildup, basically evaporation as well).
Let's take a moment to recognize what a dream "ecological cost" evaporating water is compared to old-time industries and the real environmental problems people have had to deal with. Old-timers in Cleveland can tell you how, until the 1970s (before the first serious ecological protection enforcement), the Cuyahoga River running through the city would CATCH FIRE and BURN Bible-style because of all the unprocessed, oil-based waste being dumped by plants and factories along its course. It is an unfortunate reality that many key industrial processes of our civilization dissolve dangerous and toxic compounds with water.
Also, the cost of evaporation cooling is not fundamental to data centers. You can change things around with some known engineering solutions and the costs for it would not be a deal breaker. For example, in Belgium, they built a two-loop water cooling system that can use industrial waste water (or even seawater in principle).
If you absolutely must, you can also build a fully closed-circuit liquid cooling system (think big fridge). The thing is that some water drawn from municipal system in a little city in the middle of nowhere isn't a problem.
As for high power consumption and "climate change impact," none of this is specific to data centers.
(And, this whole mindset about climate change in this podcast is just so 2010s. No, energy consumption is not inherently bad and sinful. No, the math of solving climate change with consuming less, putting on a sweater and saving does not work. A society that does not prioritize building for plentiful, cheap, (and yes,clean) energy is doomed to stagnate and wither economically. I see even most leftist people change their mind about this over the last few years, tired of never-ending green washing. If only political orthodoxies were able to change with the times...).
I was just looking at water usage in the Las Vegas area and found that one Google data center in Henderson consumes 320 million gallons a year. Just for comparison I looked at golf courses in the area and 1 golf course consumed 450 million gallons.
I see quite a few golf course near The Dalles, Oregon. (“The” is actually part of the town name).
The Dalles is just at the end of the wet side of the Columbia river gorge before it starts getting to a semi-arid region (starting about at Biggs Junction). Golf courses are more desirable toward the dryer side because it’s hard to get a game of golf in when it’s raining all the time. It is still a water limited place.
> Data centers consume a lot of water. The example they start with is Google's data center in Dalles, Oregon, which used 355 million gallons of water in 2021.
For comparison, one acre-foot of water is 325,850 gallons. Google’s data center used around 1090 acre-feet in 2021. One acre of alfalfa requires 4-6 acre-feet of water per harvest, so another way to look at it is Google’s data center used as much water as 218 acres of alfalfa. There are a million acres of alfalfa growing in California.
https://alfalfa.ucdavis.edu/sites/g/files/dgvnsk12586/files/...
I'm not sure open loop cooling is really possible in those parts of the South. The idea that water consumption on the West Coast and South East is comparable is probably flawed.
Where does the water go? Down the drain? Into the atmosphere? Broken down to component atoms?
Or you could give the key points instead of asking people to consume 2 hours of content.
I am not OP, but, sure, here are the specific facts on "ecological costs" from this podcast (with some help from Perplexity's amazing youtube summary feature):
• Data centers consume lots of water. The example they start with is that Google's data center in Dalles, Oregon used 355 million gallons of water in 2021, which was 29% of all water consumed in the city (its population is 15K though, but they neglect to mention that).
• By 2023, hyperscale data centers used 66 billion liters of water in the U.S.—triple their volume from less than a decade earlier.
• They quote estimates that ChattGPT consumes 500 mL of water for every 10-50 user prompts (or 10-50 mL per prompt).
• in Ireland datacenters in total draw "over 20% of national electricity", which outstrips the total energy usage of all urban homes in the country.
• In Cerrillos, Chile local residents blocked Google's plans to build a datacenter there after discovering the scale of water use the centers would require (169 liters per second iand its a drought-stricken area).
• Power consumption of data centers is enormous and puts a very high load on energy grids across the world.
My impression is that this is clear example of politically-biased podcast with alarmist and accusatory tone, where none of the facts presented are particularly damning in the grand scheme of things.
In data centers with large water consumption, the water mostly (90%) goes to evaporative cooling (they let hot water turn into vapor carrying away the heat) with the rest going to 10% humidification systems (getting to 40-60% humidity inside to prefect static electricity buildup).
Let's take a moment to recognize what a dream "ecological cost" it is - turning water into vapour - compared to the old timey industries and real environmental problems people have had to deal with. Old timers in Cleveland can tell you how until 1970s (before first serious ecological enforcement), Cuyahoga river running through the city would once in a while BURN WITH FIRE Bible-style from all kinds of unprocessed oil-based waste being dumped by plants and factories on its course. It's the unfortunate reality that many of our vital industrial processes that make our civilization possible rely on dissolving all kinds of most dangereous and toxic compounds in water.
Also, the cost of evaporation cooling is not something fundamental to data centers and is not something that cannot be altered with some known engineering solutions and manageble cost overheads if there is a need. For example, in Belgium they built a two-loop water cooling system that can use industry waste or even sea water. You can also get a fully closed-circuit zero running water cooling system (fridge-style) if you absolutely must.
As for high power consumption and "climate change impact", none of this is specific to data centers. We might like it or not but our society runs on energy and electric power. This whole mindset showcased in this podcast that all energy consumption is something bad and its all about reducing it is so 2010s. I think, its obvious by now that this the road to pure economical if not civilizational suicide. A society that does not prioritize building for plentiful, cheap and hopefully clean energy is doomed to wither and stagnate.
Burning gas to power this DC and cool it in a hot climate shows how hollow their net zero promises are.
If solar is cheaper per watt than gas then they will build solar and use the gas plants to level the output. Assuming the grid capacity and/or nearby land is available.
Thank you, appreciated - and the commentary as well.
People wonder why good journalism is in decline then say this kind of thing.
Audio is just one form of media, though. I prefer reading as I can consume the information more quickly (and it tends to be denser and can be more precise and detailed), and am happy to pay for quality written content.
Paris Marx is writing a book on data centers, so you might get your chance yet! If you want to read about it now tho, he does cite all his sources and transcribes the episodes.
Feed it to notebookLM and have it make a nice text file.
Good reporting would present the key parts of the article up top, but invite you to read the full story if interested. The lead paragraph and the inverted pyramid are key concepts developed in the field of journalism.
Good journalism is stating the relevant facts concisely, not blabbering for 30 minutes.
There's Ai tools for this now bro
As you might imagine, a two-hour-long podcast in a series named "Tech Won't Save Us" has time to explore many avenues of criticism.
Obviously, to summarise I have to remove the supporting examples, and the dozens of different people being interviewed. To be clear, the journalists aren't personally making all the criticisms, just interviewing other people, so if some of the following seems to contradict itself, that's why.
A fair chunk of the podcast involves explaining the context to a broad audience. You know, explaining what a data centre is, outlining the cloud market and its major players, etc.
The criticisms outlined in the podcast include:
* Data centres produce very few jobs for the communities they're located in.
* They are often built in struggling communities where 'enterprise zones' offer big tax breaks, hoping to attract employers.
* They consume quite a lot of power - not as much as, say, an aluminium smelter, but perhaps as much as 150,000 homes. Few cities have that much spare grid capacity, and some have warned about risks of rolling blackouts.
* 20 percent of Ireland's electricity is used for data centres (they're something of europe's data centre capital due to their attractive tax rates)
* Energy demand at data centres leads to greater emissions at power plants. Even if the data centre contracts to only buy renewable power, that might displace less-eco-friendly buyers of renewables onto non-renewable power sources. And a lot of things like 'carbon credits' are based on rather creative accounting.
* One high-profile data centre (in The Dalles, Oregon) is in a town suffering a drought, and consumes quite a lot of water, considering it's a drought area. Grass on the local golf course is completely dead.
* Land and tax breaks are often acquired through secretive shell companies that insist on secrecy agreements with desperate local governments; in one case the government didn't even know they were dealing with Google. This secrecy extends to agreements about things like water usage.
* As you can imagine, a local community suffering a drought sees the local data centre's water consumption being kept secret by elected officials, they assume the worst.
* Some data centre builders, like Elon Musk, have a history of making legally non-binding promises, then not bothering to keep them. And of running large gas generators without permits.
* The kind of distressed post-industrial communities that welcome data centres often have high levels of pollution and cancer, making those unpermitted generators particularly bad.
* Many of the hyperscalers are also big AI boosters, so it's not like the datacentre operators can disclaim responsibility for the power needs of AI.
* Many people have criticisms of AI, beyond energy consumption. Such as huge centralised LLMs transferring more control to huge tech firms; getting things wrong; AI friends being an alienating concept; having heavy-handed censorship; widespread use of bots on platforms like twitter and reddit; risks of job losses; being trained on pirated ebooks without authors' permission; being a really shitty therapist; producing mediocre art; producing porn depicting real people without their consent; producing creepy underage porn.
* Or AI might be a bubble that's about to burst, which would also be bad but for other reasons.
* Tech business leaders like Sam Altman are on record saying some pretty wacky things about AI power consumption, like that the high power demands of AI will force us to invent fusion power. A load of them also have weird, messianic ideas about "the singularity", or think we're all in a simulation already, or think living in a Matrix-style simulated 'metaverse' sounds like a great thing.
* Many of the highest-profile tech folks - the billionaires - have very right-wing politics. Such as opposing all regulation as a matter of principle, except on the occasions when it works for their benefit. Some people think expecting these folks to regulate themselves isn't the best idea.
Overall this is all stuff that followers of tech industry news will probably have heard before; the podcast just adds context, draws it together, and finds sources in the form of interviewees.
And yet computing power is what we need to advance technologicaly. Unstoppable force that indeed asks for solutions to new problems. Better acknowledge them then to fight them.
I feel like we need more social equality to advance technologically rather than gadgets. The idea that only the elites and rich can conjure technological wonders is just so demonstrably false and not needed in this moment of history.
Uplifting everyone ensures that we'll be that much more likely to find the next Mozart or Tesla or Torvalds or whoever, if we give them a chance.
But yes, better to acknowledge how capital can be better utilized. You can probably give away free school lunches for an entire generation of children with that $10 billion in Louisiana, or you can give it to Zuckerberg to get slightly richer.
Becomes abundantly clear which one is better for societal advancement.
> You can probably give away free school lunches for an entire generation of children with that $10 billion in Louisiana
The annual cost of the National School Lunch Program is $18B, so, no.
Do you want one in your city? Taking water and power from your power infrastructure? I don’t, and thankfully our city government rejected an Amazon proposal to do that.
From one perspective, if they don't create many jobs and don't pay any taxes, I can see why communities don't welcome them.
From another perspective, compared to a coal mine or a paint factory or a steelworks or an airport or a landfill or an oil refinery, data centres are safe, low-pollution, low-noise, low-odour, and low-traffic. By the standards of industrial areas, they're great neighbours.
Great podcast, highly recommend it. Paris Marx is writing a book now on data centers. Between him and This Machine Kills it feels great to finally find some high quality tech journalism.
404, wired, and the verge have been doing good work recently too.
Access to very cheap power in the MISO region is likely one of the top driving factors for this location. It extends partially into Texas and I've found that my rates are sometimes as little as half of what ERCOT customers are paying.
The #1 thing that makes MISO so cheap is the fact that it has the heaviest coal generation mix (>40%) out of all US regional grid operators. Any talk about natural gas or renewables pales in comparison.
I would guess the 20 year 'tax break' (AKA the other taxpayers are footing the bill) is the real reason for the building. shell game
Meta built a data center in North Kansas City. I'm not sure details of their break (Mayor loves to hand out money), but power is likely cheaper, & def much greener (1/3 from wind farms in Western Kansas state last I checked).
"take advantage of a new Louisiana incentive program, established by Act 730, that offers qualifying projects a state and local sales and use tax rebate on the purchase or lease of data center equipment"
https://www.opportunitylouisiana.gov/news/meta-selects-north...
> AKA the other taxpayers are footing the bill
Curious as to how you reached this conclusion. No taxpayer funds are going towards construction or operation of the data center. The lack of tax revenue from Meta is nothing spent, and they're still going to be paying into the local economy. The energy infrastructure is going to be built by Entergy, who've projected it to cost customers ~$1 more or less per month.
As someone who lives here, this is one of the few times I agree with our government. We're one of the least competitive states in the country, our tech sector is almost non-existent. It's reasonable to offer what you can to attract business. I think Landry's LED efforts so far have been a respectable attempt at improving the state of things.
Why would coal generation make it so cheap? I would expect any coal-heavy generation region to be far more expensive than a hydro-heavy region.
https://en.m.wikipedia.org/wiki/Cost_of_electricity_by_sourc...
Because all of the plants are fully amortized. If they were brand new, the total cost would be excessive. Customers are mostly just paying for operations, fuel and maintenance. There is no capital recovery needed.
That's true of hydro heavy regions too, if not even more so. The Hoover Dam is not new. The Bonneville Dam is not new. Niagara Falls is not new.
Not that it changes your point that much, but doesn’t MISO have more like a 20-30% Coal mix?[0]
It looks like natural gas is usually the biggest source of electricity.
[0] https://app.electricitymaps.com/zone/US-MIDW-MISO/72h/hourly...
If you look at the all years view on electricity maps (https://app.electricitymaps.com/zone/US-MIDW-MISO/all/yearly), 25% is share of generation from coal in MISO.
I'd be curious to test the GP's point. Since electricity maps doesn't have cost data for most US balancing authorities, you maybe could try figuring out power costs per balancing authority to end customers by using something like the https://www.eia.gov/electricity/data/eia861/ "Sales to Customers Customer Sited" data. Revenue over Megawatthours for Industrial service.
You can look on the real time map:
https://miso.singularity.energy/realtime
I am seeing 39.1% right now
Location is in Holly Ridge, LA.
It is bounded by Fortenberry Rd on the N, LA183 on the E, US80 on the S and Jaggers Ln on the W. It overlays Burn, Wade and Smalling roads.
https://www.richlandparishdatacenter.com/blank-5
It has 6 reviews and a 3.7 rating on Google. https://maps.app.goo.gl/pxXR5zxfiiBDDNrB7
construction website: https://www.richlandparishdatacenter.com/
It's a Zone X flood zone bounded by Zone AE that runs through the parcel. 22083C0260D
It's a pretty piss poor location to invest a boat load of money without putting servers in actual boats.
Just make it floodproof and get free cooling!
They need IPX67 servers that can double as submarine thrusters.
Louisiana is so historically corrupt : meta chose Louisiana because they can pretty much do what they want and the parishes (counties) like Richland where this is being built are so poor they couldnt care less about any environmental or quality of life setbacks. as a reminder , Louisiana is home to the cancer alley.https://en.m.wikipedia.org/wiki/Cancer_Alley
Reliability is king. Why not just buy 50 sq mi of ranch land in Arizona for a data center complex, install a solar and battery storage system and eliminate electricity concerns for the next 20 years? Using LFP or sodium-ion batteries for storage effectively eliminates fire risk.
Have there been any formal studies looking at OpEx ROI for offgrid carbon free generation + storage for data centers? Exiting the grid and vertically integrating on site generation eliminates a lot of risk when dealing with an external utility.
FAANG has the market cap to drive down costs for reliable carbon-free generation and storage, like Alphabet is doing[1].
Microsoft is demonstrating water-free cooling solutions[2]. As long as there's a fiber backbone nearby, FAANG can slash energy OpEx and not worry about the rest of the grid. Or natural gas prices.
[1]https://dataconomy.com/2024/12/11/why-googles-800m-bet-on-cl...
[2]https://datacentremagazine.com/news/how-are-companies-pionee...
While Meta has a non-binding promise to build more renewable energy, the Louisiana Legislature passed a new law that adds natural gas to the definition of green energy, allowing Zuckerberg and others to count Entergy’s gas turbines as “green.”
As much as I prefer burning gas over coal, conflating it with zero(-ish) emission energy sources like wind, solar, and nuclear is bad.
Due to all the methane leaks, gas isn't even as much cleaner than coal as it was purported to be... But hey monitoring programs got cut so I guess that solves the problem...
From a purely greenhouse gas accounting, sure.
Anyone who has to live in a fairly closed system (i.e. this planet) in which fossil fuels are burned for power would be beyond a fool to not strongly prefer gas over coal seeing as their greenhouse emissions are close enough to be within arguing distance. It's all the other stuff coming out that's the problem with coal.
I think you might have a typo. Reading your comment literally, it doesn't make sense.
Summarized: Anyone would be a fool not to prefer gas or coal, because their emissions are nearly equal.
One doesn't follow from the other, can you correct/elaborate?
I think the point is: "you'd be a fool not to prefer gas, because while the greenhouse emissions are about the same, for everything else coal is much worse"
They said gas over coal. If you accept the claim that GHG emissions from gas and coal are roughly equal, their claim is the other pollutants from burning coal make gas far more preferable.
If their greenhouse emissions are even close only a moron would not pick gas over coal because the former's emissions lack all the other nasty byproducts that are present in the latter's emissions.
I agree methane leaks (and monitoring programs cuts) are a problem. But even with them, methane burns much more cleanly than coal. The former primarily emits CO2 and H2O, while the latter emits SO2, NOx, heavy metals and more.
These definitions always get muddled when flipping between CO2 emissions or pollution... coal is definitely worse from a pollution standpoint, is likely worse from a carbon standpoint, but much of the methane produced from natural gas production is just released into the atmosphere and has a dramatically higher warming effect compared to CO2 -- on the order of 80x more warming potential over 20 years and at least 20x over 100 years.
So only looking at the byproducts of methane combustion is also misleading since nat. gas plants largely aren't burning methane - and blanket statements for all natural gas are also misleading since e.g. the gas from Canada is extremely 'Sour' and releases a ton of sulfur compounds when burned, often with fewer scrubbers than coal plants.
This is a really interesting comment. Do you have a reference for the 80x figure, or the “sour” Canadian gas? Would love to read more about this
Methane mostly disassembles into CO2 but it takes 12+ years. When thinking about global warming potential, everything is compared to CO2 which we’ve normalized as “1”. So something with a GWP of 2 is twice as bad as CO2 in equal volumes.
Methane will eventually break down into CO2, so if you look at the GWP for years 13-100, it’s 1. The weighted average for years 1-100 is over 20x, so it follows that if you look only at a shorter time frame, it would be dramatically higher and is indeed - somewhere north or 80 for a 20-year time frame.
https://www.epa.gov/ghgemissions/understanding-global-warmin...
As far as sour gas is concerned - not all natural gas formations are created equal. If you look at any serious pollution evaluation, they take into account which formation the gas was harvested from. Texas gas is pretty ‘sweet’ with low sulfur and acid content but much of the oil/gas in Western Canada or the Gulf is ‘sour’ and must be treated and refined prior to being sold as fuel. So it also follows here that flaring methane from sour fields is going to release a bunch of the souring compounds and have a much stronger environmental impact as compared to sweet formations.
https://nsrp.vn/latest-article/sour-crude-oil-and-sweet-crud...
I think the problem is that methane is 20x more powerful a GHG than CO2
Laugh in the face of anyone suggesting CO2 capture technology. We won't even capture the more-valuable methane.
[dead]
As an aside, methane leaks from coal mines can be worse than upstream leaks from O&G.
burning fossil fuel and depleting the local water aquifer, I'm starting to miss the greenwashing era
Behaving a certain way to pretend being virtuous, it turns out, is almost as good as actually being virtuous.
Is there really a concern that the datacenter is going to drink up all the water in Louisiana?
I was much more concerned that it will be expensive to cool because it's situated in a state with a lot of hot and humid days.
Redefining words to fit their narrative and premise...hmm where have I seen that before?
Who is this non-binding promise being made to, and why make one?
"I'll gladly pay you Tuesday for a hamburger today…" Seems to be pretty common these days when corporate make deals with cities/counties/states.
[dead]
[flagged]
[flagged]
> Meta has a non-binding promise to build more renewable energy
Also the people working for that company. Unimaginable wealth, both at the corporate and personal level, everyone aware at this point that the climate is breaking down and yet, they just can't do the right thing because they are just too damn greedy.
Adding natural gas to the definition of green energy is absolutely wild. How on earth did that pass?
Louisiana has a long history of political corruption, and the petrochemical industry is a major part of their economy.
LA has the resource curse.
I have to imagine it's just a complete lack of care and classifying it as "green" helps push through something that they're being lobbied to push. I can't imagine this is anything but nonsense.
We all know how it passed. Legislators have lots of money in natural gas I’m sure.
Looks like Louisiana is all aboard the "internal colonialism" that seems to be all the rage at the state level lately. In this case, flouting national/international renewable energy policy so the good people of Louisiana can get the long term benefit of... Having to deal with the fallout of another datacentre project?
Come on Louisiana legislature, at least make them pay for resurfacing a highway or something.
> Having to deal with the fallout of another datacentre project?
I don't understand. What are the specific risks facing the people of Louisiana?
[flagged]
None of those energy source is zero-ish. They all require upfront releases of CO2 to create, and end of life release to recycle.
Nuclear for base load and gas for peak/flexible demand is the most climate friendly solution available.
Look, I love to be pedantic as much as the next person on this site, but let's not miss the forest for the trees. State level legislature relabeling fossil fuels so they count as "green" is not the path to a better future.
> They all require upfront releases of CO2 to create, and end of life release to recycle.
All of them require that; but not all of them require it during the production. Some, like natural gas, do.
Yes, but that's not the point is it. The point is: what are total emissions?
I'm not super-familiar with Louisiana, but my general impression is there's a lot of climate/weather events that are gonna impact power reliability. Hmmm.
If you're thinking of hurricanes (and you may not be), the location is far enough away from the coast that they wouldn't be a significant problem.
If you've not paid attention to the recent hurricane damages to the US, it wasn't just coastal cities that were hammered. Lots of places "far enough away from the coast" saw lots of flooding. A hurricane doesn't just evaporate. The hurricane reverses the process back to Tropical Storm, Depression, etc while continuing to bring lots of rain minus all that wind
I'm assuming you're thinking about Helene. That was a very unusual situation that had a lot to do with local topography and geology on top of a very unusual weather situation (ground-saturating rains in the week prior to a hurricane).
This area hasn't been underwater in a meaningful way since the 1927 flood of the Mississippi.
Do you think these 100 year events are going to occur more or less frequently now?
Does it matter what I think?
Any situation that floods that area is going to be well beyond 100-year floods, likely beyond 1000-year, and (like the 1927 flood) predicated on catastrophic levee failures that result in the Mississippi being over 100 miles wide.
Is it possible? Of course. Is it likely? No. Would there be much bigger issues than a server farm being offline? Yes.
The worst-case scenario from a storm that hits locally (which, e.g., the 1927 flood was absolutely not a consequence of - it was water from upstream, which climate change will weaken because of decreased snowmelt) is like Harvey over Houston (similar terrain). But this is much farther inland and could not replenish itself from the Gulf.
That said, it is in the middle of nowhere.
But why would we think Meta hadn’t assessed that particular risk?
Looks like it's surrounded by ponds to contain potential flooding. And it's apparently getting 3 new power plants.
> While Meta has a non-binding promise to build more renewable energy, the Louisiana Legislature passed a new law that adds natural gas to the definition of green energy, allowing Zuckerberg and others to count Entergy’s gas turbines as “green.”
This means it isn't securities fraud when Meta tries to meet "climate commitments" due to the greenwashing of fossil gas generation by the state of Louisiana. Louisiana is a low regulation jurisdiction that doesn't care if most of the state ends up a Superfund site, so it is ideal to colocate data centers that are going to burn up a bunch of fossil gas there over their lifetime (when they are unwelcome elsewhere).
https://en.wikipedia.org/wiki/Cancer_Alley
https://www.propublica.org/article/toxmap-poison-in-the-air
https://www.propublica.org/article/cancer-alley-louisiana-ep...
https://www.propublica.org/article/welcome-to-cancer-alley-w...
While I like solar & wind, when oil is drilled, natural gas is often burned off at the drill site. So if it is going to be burned, might as well make electrify from it.
"World Bank is urging energy firms to gather the gas and sell it to businesses and consumers.... Companies can use the gas in mobile electricity generating stations, to power their oil drilling sites, or as a fuel in petrochemical plants."
https://www.bbc.com/news/science-environment-63051458
https://www.worldbank.org/en/programs/gasflaringreduction/ga...
https://www.reddit.com/r/explainlikeimfive/comments/cvawx7/e...
As long as it is fossil gas that would've otherwise been flared or not captured at all, I strongly agree.
And people say blood sacrifices have ended in the modern world.
A major Mississippi River flood could drown the site, but generally I’d think the risk level is fairly low.
It’s north of the Old River/Morganza flood control system. I think this would have to fail first in order for the Meta site to be in danger.
Also, if the New Madrid fault went off again, I think you’d kiss this site goodbye. But if that happened you’d have about a million higher priority concerns than a stupid data center.
Far enough away from the coast... so far.
Most DCs have SLAs with energy companies and have redundant sources from independent plants, not to mention generators and batteries.
Interesting that they bring up water consumption https://www.bbc.com/news/articles/cy8gy7lv448o
I am amazed that it appears that Meta did not ask for tax breaks for a $10B project. Seems like they absolutely could have 'bid' this out among competing locales.
They didn't have to ask for special new breaks; the breaks were already passed:
https://www.opportunitylouisiana.gov/news/meta-selects-north...
Thank you looks like they did get millions a year in no sales or use tax on data center equipment. Nothing compared to what NFL teams con out of cities for stadiums.
There is a big matrix of risk/reward for any DC location.
You bet Meta asked for incentives, but sometimes a guarantee of future power capacity, fast permitting, or ideal locations are worth more than the incentives the state could afford.
Sure, but you don't have to build your DC at the place that has the cheapest bid.
You just need to make them think you might not build somewhere else unless they sweeten the deal.
Hopefully this just means that governments have wisened up to the fact that a gazillion DCs are going to be built so if you pass on Meta you can just pickup Google's.
No. Meta is spending $10B to build its largest data center in rural Louisiana.
I guarantee you that a lot of that $10B will be spent out of state. This is yet another corporate handout with the thin veil of "technology investment" Louisiana loves.
Myth: It has computers, therefore we are investing in technology and hi-tech jobs.
Fact: This will be built by out of state contractors, staffed by mostly out of state workers, and far less than anyone expects or claims. And will essentially transfer local resources out of the state while making those resources more scarce for residents.
Louisiana let EA run their QA from here to severely underpay people and pay fewer taxes. They courted IBM to do the same with Salesforce jobs. And now Meta gets to exploit the state to enrich another out-of-state corporation.
I always told myself if I ever became a "tech billionaire" I'd buy out a random abandoned town somewhere, setup high speed internet, and turn a ghost town into a high tech town, cause why not? You could easily become mayor and approve some reasonable projects. Sell extremely affordable housing for the buck (close to actual cost).
I do often wonder if it might be worthwhile to shove a bunch of server farms into a few abandoned mines, if you setup the appropriate infrastructure in said mines to protect your data centers.
I always thought Detroit would have been the ideal location for Amazon to build HQ2 for that reason.
https://en.wikipedia.org/wiki/Fordlandia
You can build it, sure, but why would anyone want to live there?
Disconnected from a major city, fiber internet, affordable, why would anyone want to move to Orlando it's all swamp lands? I remember that being Walt Disneys response when a reporter (who was correct) accused him of buying up all the land in Florida.
97% of the USA is already disconnected from a major city and affordable. And now with Starlink broadband internet is accessible as well. Yet there is no rush to migrate to rural areas - quite the opposite. People want to live in cities and suburbs.
Sure, but most people don't want to live far from a city with limited resources like access to doctors, and such.
> affordable
You'd get Waco'ed or Ruby Ridge'ed.
What could these data-centers used for post-bubble?
For BTC and high tech surveillance
Even if the LLM bubble pops we could still do other AI things on the overcapacity. It'd be awesome renting a B200 for $1/hr or something.
The major tech companies are all scrambling to snap up cheap energy right now. The result is that we are dumping a whole lot of additional carbon in red states and adding a while lot of additional extremely expensive per MWh sources in blue states. In both cases, the winners will be tech company shareholders and the losers will be the people who actually live in these communities who will end up with dirtier, more expensive power.
The losers are going to be the energy companies who think they’re getting long-term energy sales but probably won’t be, since these techniques will get more efficient.
https://en.wikipedia.org/wiki/Jevons_paradox
People using consumer generative AIs are already using it for free, or very cheaply. It may be hard for falling costs to drive more demand.
The techniques will get more efficient, but the quantity of training will increase monotonically. We aren't going to use less energy overall. The ratepayers are absolutely the ones who will lose out on this.
The demand for energy will never go down, the more we can produce the more we will use.
The article says
> Electricity demand in the U.S. held steady for 15 years but, last year, it increased by 3%— marking the fifth-highest rise this century. More jumps are projected for years to come.
https://www.eia.gov/energyexplained/electricity/electricity-...
Total electricity generated has been relatively flat for a couple decades.
Surely EVs would change the equation. Also increasing installation of heat pumps vs gas heating.
Perhaps, my point was an electricity supplier that invested a ton of money in the early 2000s assuming that aggregate demand will keep growing forever would have been in for a rough time.
A variety of factors may or may not make a future where aggregate electricity demand would increases, or stagnates, or even declined.
Translation: Meta is wasting their money and destroying an already struggling community and the environment to be participating at something at which they are both hopelessly late to and hopelessly behind in.
Can't wait for their stock to crash and Zuck to be out.
With zuck’s dual class stock he will never be removed. He will either leave on his own or the company will no longer exist.
I am fine with any of these options. I have my fingers crossed for an Adam Neumann kind of exit.
These things destroy local communities. But none of the locals have access to the fuhr
I'll never understand why tech companies choose some of the locations for their data centers. Considering a big thing with data centers is "keeping stuff cool", you would think they would build them in the northern states, closer to Canada versus the hot sticky swamp.
> I'll never understand why tech companies choose some of the locations
That's because you've chosen not to read about it. Location is one of the most important things they think about for data centers and there are plenty of articles on the subject.
Here's a recent article:
https://www.cnbc.com/2025/06/25/meta-massive-data-center-lou...
“We set out looking for a place where we could expand into gigawatts pretty quickly, and really get moving within that community on a large plot of land very quickly,” said Rachel Peterson, vice president of data centers for Meta. “We looked at finding very, very large contiguous plots of land that had access to the infrastructure that we need, the energy that we needed, and could move very, very quickly for us.”
To answer the question you're implying, surrounding temperature is pretty minor, the cooling required is orders of magnitude higher, so power access is more important; You'll frequently find them located near sources of energy.
Meta has defacto infinite money, they don't have to look for places where operation is cheap, but where they can be above the law as much as possible for doing whatever they want.
I'm pretty sure Meta's team has written about that at length. It's about many things, such as (power/transportation/internet/energy) infrastructure, political situation, available workforce, vicinity to population centers, property prices, and a whole lot more
It says right in the article that they have lots of natural gas, and the state is bringing on 2GW of new electrical capacity.
Cheap gas, cheap land (it's on a big essentially empty plot that people have wanted to develop for a while, in a poor area with plenty of underutilized farmland), state and local governments that care more about this project than about environmental concerns.
Similar reason to why a lot of chemical manufacturing is in Louisiana.
Cheap land and cheap energy.
Louisiana and New Orleans have been pushing to make the city a “tech hub” for the past 10 years (why would you build data centers in a flood-prone basin below sea level? I don’t know). I imagine most of it is striking a sweet heart deal with the municipalities that want the business.
There were several data centers and colos on Poydras St in downtown when Katrina hit. Famously SomethingAwful was being hosted out of one of them whose remaining on site employee live blogged the whole thing on LiveJournal.
This will be built in North Louisiana from what I understand, well above sea level.
Hurricanes on the other hand will still be a very real thing.
The speed of light is incredibly slow and data through a wire is even slower. Proximity is worth something
>The speed of light is incredibly slow
I get where you're coming from but still I find funny in so many levels that the literal speed limit of the universe is too slow for our mundane (or even banal in FB case) needs. the universe isn't good enough to our need to move bullshit across the globe. surreal.
In the same vein it would be awesome if this _need for speed_ would materialize in infinite funding of neutrino based communication research.
As a side note, if you liked the above comment, but haven't yet read "A Fire Upon the Deep", you will probably enjoy it.
Doesn’t matter for training, as all GPUs are colocated in the same DC.
Distance is a minor factor. They'd put a data center on the moon if it had an abundance of cheap energy.
So why put a datacenter in Louisiana, far from the vast majority of people in the Americas?
You’re pretty close to Texas tech hubs, plus Meta was able to convince them to pass Louisiana Act No 730 so they save a ton on capex
> The project entails more than 2 gigawatts of computing capacity—Zuckerberg said it could eventually expand to 5 gigawatts—programmed to train open-source large language models.
Given that the human brain takes much longer to "train", I wonder how the energy efficiency pans out — comparing the two.
How long does a human brain take to train?
Biological systems are wildly energy efficient, that's kind of their whole thing. The average human will consume approximately 75kwh worth of calories in their lifetime. There are electric cars with bigger batteries.
[Edit] ok, yes, please. I get that i missed the k in kcal. The point stands. Biological training is massively more efficient, even when you forget to multiply by 1000
This is wrong by at least three orders of magnitude. Very roughly, a human requires 2000 kcal a day = 2 kWh a day so 75 kWh is enough to cover about a month, putting aside the upstream losses in the energy supply chain (which are far greater for humans).
In general, saying that biological systems are "wildly efficient" is... wildly wrong. Some biological processes are optimized by evolution... most are not. There are no bicycles in nature.
> The average human will consume approximately 75kwh worth of calories in their lifetime. There are electric cars with bigger batteries.
Doesn't pass the smell test. I think I could push an electric car at least a mile a day if that's what I spent most my extra calories on. If I did that I'd surpass its range in well under 2 years, much less than my lifetime.
You're off by about three orders of magnitude.
A human consuming 2000 kcal/day (conservative estimate) uses about 2.32 kWh per day. Over 75 years, that's roughly 64,000 kWh.
Oh, right i did a conversion wrong. Woops. In any case, a rounding error when talking about gigawatts of generation capacity
We're efficient once we have the energy, sure. How much energy does it take to go from raw sunlight to a calorie your body is actually able to use, and finally to your dinner table?
All of our food was alive before we ate it. All calories used by living things are efficient. Life is an end unto itself. It does not need to justify its existence by the moral code of technocrat materialism. The fact that this discussion is being had on this board in good faith is morally condemning of our worldview.
Since the original point of this chain was a comparison between the energy efficiency of biological vs machine learning, then we need to be trying to understand if the machine is more efficient than the human. You don't need to make some moral or philosophical argument about existential justification to accept that taking a more efficient approach is better, in that it generally enables more life for the same energy.
If the true, total cost of a machine to perform some task is less than a person to do the same task, then the machine should do it and the person should move to do what the machine cannot. This means more energy is available for everything else, living included.
Your forget that a biological system has approximately 0 throughput in work done.
Nearly everything a biological system accomplishes depends on massive external machinery.
Humans are only intellectually interesting because of their use of tools.