45°C cooling design cuts data center water use to near zero
blogs.nvidia.com
blogs.nvidia.com
Summer is still an issue, but fun solutions are possible. With the right geology, I think it’s possible to heat an underground volume in the summer and recapture (some of) that heat in the winter. In many, many climates, annual heating costs are far higher than cooling costs, at least if people aren’t stupid with skylights. [0]
[0] As a back-of-the-envelope heuristic, heating or cooling load due to conduction and air exchange is proportional to the difference between indoor and outdoor temperature. Outdoor temperatures of -10F to 30F are not unusual in the winter and are 40-80F away from an indoor temp of 70F. But outdoor temperatures in these climates rarely exceed 95F and are mostly lower in the summer, so that’s 15-25F of cooling. And heat pumps are more efficient at smaller temperature differences.
Radiative heating is an entirely different story.
(Seasonal heat storage is also a thing, Espoo's neighbours have tens of GWh of storage, with a new 90 GWh cavern in the works. Not sure if the systems are interlinked.)
Apparently the local supercomputer cluster in Kajaani has also been hooked up in 2021 and is responsible for a fifth of heating.
Also interesting that the article states that this engineering problem hadn’t been solved before. Google pioneered running chips hotter than before. Moreover, we have had water cooling in consumer setups for ages. (At least 30 years.) So what is new is that all chips have been attached to the loop. I couldn’t find what they did with PSU though.
A 75-MW data center in Mäntsälä has provided 2/3 of the heat for the town (2,500 homes) for a decade [2]
1. https://www.creatingsustainablecities.org.uk/post/case-study...
2. https://www.sustainabilitymenews.com/waste-management/how-fi...
Then 45 or below is sent back on the return.
DC inlet is 45°C, outlet is 55°C assuming a 10°C ΔT. By the time that's travelled 500m–1km through pipework you've lost a few degrees, so you're arriving at the HIU at maybe 50–52°C. The home radiator circuit then takes that down by around say 12°C, returning ~38°C. Factor in pipe losses on the return leg and you're back at the data centre with maybe 35°C inlet rather than 45°C — meaning the DC output is now only 45°C rather than 55°C, and the whole system gradually degrades each cycle. You could address this by mixing some hot output back into the return to keep the DC inlet stable at 45°C, but eh.
you'd be surprised. If you have high flow, and the pipes are insulated and underground, then after about a week the temperature drop isn't that much. You do have heat losses, but if you have a high enough continuos flow and a big enough pipe, then it'll be low enough to not worry about, especially if the aim is heat shedding rather than efficiency.
My old flat was powered by both the 1970s boilers across the way, and more recently the massive south london incinerator. The pipe cross section was I think 40cm and at peak carried ~3-5 Megawatts of heat. I think it operated at 150c, but that could be me misremembering (this is the later version of the network: https://www.burohappold.com/projects/veolia-southwark-2-0-he... for the councils is they get a maintained heater network, which is much cheaper than doing it themselves (even more now with gas being so expensive) the power station gets to charge for a waste product and it doubles their on paper efficiency, its a win win.)
Surely having the input fluid being colder is a benefit, not a problem? Just run the fluid more slowly through the system?
*55C is on the output, read article first pls
Main problem is that it wouldn't work with buildings designed for higher heating temperature so it is limited to new builds. And it is not limited just to replacing heaters, hot water system is also designed to work with higher temperatures so heat exchanger used would have to be significantly larger
Another one is that load is not constant on both sides and not exactly something that can be increased on demand (unless you're fine with burning cycles just to electrically heat, but that's massively inefficient)
That's no longer the case with modern, multi-stage high temperature heat pumps [1]. These consume more electricity, yes, but still achieve far better efficiency than straight burning fuel or resistive heaters.
(seriously though, I knew they had hot water, it just never occurred to me how awesome it is to have hot water on the tap)
Only about 2% of the total heating used though...
It's got a "heat energy to/from campus" exchange in there.
That's a link in March and the air temperature was 31°F.
https://web.archive.org/web/20210708150410/https://www.nrel.... is later with air temperature of 68°F.
https://web.archive.org/web/20210708150410/https://www.nrel....
The thermosyphon took it from 90.2°F to 70.8°F ... and then it went to the counterflow cooling towers which took it from 69.8°F to 64.8°F. If that thermosyphon did the majority of the work and there's not too much more in the cooling towers... and that was only rejecting 375kW of heat through evaporation.
Note one of the parameters with it in the dashboard: dry bulb temperature.
https://web.archive.org/web/20251207023030/https://docs.nrel... for more info on the thermosyphon.
I'm disappointed that the dashboard isn't available anymore. It was rather neat to look at.
Some older articles about it:
https://www.nlr.gov/news/detail/program/2023/a-decade-of-gre...
https://www.nlr.gov/computational-science/data-center-coolin... (click through these)
When possible, heat energy from the energy recovery loop is transferred to the building process hot water loop, which provides heat for the office and laboratory spaces within the building. The energy recovery heat exchanger (6) transfers heat from the ERW loop to the process hot water loop.
After re-use potential is exhausted, warm ERW water flows to the fourth floor mechanical room. When temperatures permit, heat is dissipated through a thermosyphon (7), which is an advanced dry cooler that uses refrigerant in a passive cycle to dissipate heat.
Remaining heat is transferred from the ERW loop to a tower water open loop via the cooling tower heat exchanger (8). Cooling towers (9) cool the tower water loop by cascading that water across fill material while drawing ambient air across the fill material. This provides a very energy-efficient way to cool water with sensible (heat dissipated to air without evaporation) and latent (heat dissipated with evaporation) heat transfer. However, this evaporative cooling process requires a continuous source of water.
https://www.nlr.gov/computational-science/reducing-water-usa...
In August 2016, a prototype thermosyphon cooler—an advanced dry cooler that uses refrigerant in a passive cycle to dissipate heat—was installed. The thermosyphon was placed upstream of the HPC Data Center cooling towers at the ESIF to create a hybrid cooling system. The system coordinates the operation for optimum water and operating cost efficiency—using wet cooling when it's hot and dry cooling when it's not.The problem is really how much energy is actually available.
Skylights also suck because they're exposing the window directly to the sky (and thus rain), whereas conventional windows are somewhat protected by eaves and the fact that the window is vertical so water can't pool on it.
Introducing skylights allows the total fenestration area (windows plus skylights) to be reduced from a maximum 20% of floor area to as low as 12% of floor area while achieving the same baseline average daylight factor target of 5%, and reduces annual heating and cooling energy use and costs in all but two of the 108 models with skylights analyzed. In other words, when different combinations of skylights and windows are used to achieve the same target daylight factor, the heating and cooling energy cost savings are almost always greater when equivalent daylight comes from top‐lighting (skylights) rather than side‐lighting (windows).
Property values have consistently gone up in that region for decades, and are up to $6 million an acre if there's enough contiguous land to put another data center on.
Many of the people complaining about datacenters would also complain about literally any kind of development.
The problem is that apparently you can just ignore that by building in poor places that won't hold you accountable and perhaps don't even have the regulations. If they do, just don't comply with them. Then your gas turbines can be as loud as you like, nobody will stop you. There's this one weird trick where you can pretend your generators aren't turned on, but they are, and they pollute badly. Nobody will call you out on it.
https://www.youtube.com/watch?v=3VJT2JeDCyw
Also you can take people's drinking water to cool your data centre, and promise a water recycling plant, and then just not build it.
https://www.politico.com/news/2026/05/05/xai-water-reuse-pro...
What I forgot to mention is this one weird trick only works if you're a turbo-bastard with an enormous bank account, and the government of the poor place thinks it's in line for a payday.
Power plants are all over, even in populated areas. They’re not so bad either (except perhaps coal).
There is no fundamental reason that datacenters need to be especially unpleasant to their neighbors.
Sure there is, being a good neighbor costs more than being a bad neighbor
I don't have any datacenters near me but I can hear some heavy hums from the washing machine 3 floors up when it put my head on my pillow, for some reason it just propagates through the building physically. When I walk around I don't hear it. Datacenter noise can be the same.
IMO they should be put away from habitation, there's no reason for them to be near there anyway
My GPUs at Hurricane Electric in Fremont are also completely unnoticeable outside the building. Inside, when I'm working at the cabinets it's obviously deafening. Outside you wouldn't even know. Realistically, the predominant sounds at my home are from the traffic on the Bay Bridge so it's nice when there's congestion because it's quiet.
Honestly, I wish there were more urban datacenters. It's getting quite annoying having to make a 1 hr trek to Fremont every time I want to rack a new server.
There's a lot of them in high rise buildings... but they come with high rise rent.
No one bats an eye when an air conditioner runs.
I find their noise pretty obnoxious. Out of respect for my neighbours, when I get around to installing one, I'll be getting the absolute quietest model available.
In the US
I don't live next to one but I'd take constant humming over the constant stop/go traffic noise, honking, squeaky brakes, slamming doors and revving engines I now have on my western side of the apartment, thanks to the unemployment office the city opened on my street not too long ago.
So how come constant humming is somehow an illegal nuisance, but we've been expected to put up with the much more annoying urban traffic noise for decades just fine?
My parents apartment have constant humming anyway thanks to the HVAC system on the roof of the nearby supermarket and white/brown noise is far more tolerable and easy to tune out than traffic noises.
For one, there tends to be little traffic at night when most people want quiet in order to sleep. Driving is also something (nearly) everyone does and benefits directly from, so negative externalities are easier to accept. It is much harder to accept a new source of noise near your home you haven't asked for and don't directly benefit from.
With that said, by the standards of industrial sites data centres are quiet, low traffic and smell free. An industrial area that can’t build a data centre certainly can’t build a steelworks or oil refinery or leather tannery.
I think we are going to need heating.
Nearly 10x more people die from the cold than from the heat.
"...9.43% of global deaths were attributable to non-optimal temperatures, with 8.52% from cold and 0.91% from heat."
https://www.thelancet.com/journals/lanplh/article/PIIS2542-5...
I don't think this is climate change propaganda, but your application of this study by evoking it in a discussion about climate change feels like it.
So what, winters would be no more? Snow will disappear, no more ice-men and christmas trees, and subzero conditions in general, too?
You do eat, don't you?
You are not wrong, but the whole issue is a bit silly: there should be legal ways for data centres (and other commercial operations) to just send a few million dollars a year to whichever community they need to convince; instead of having to dress it up as free heating.
And when it does, people should obviously go for it. Work out a deal, when there's some surplus to share.
The heat is waste heat. If it cannot be recovered as a profitable source of energy, the datacenter won't be able to pay that few million dollars.
What makes you think so? Datacentres are already very expensive, and getting permissions quicker (or at all) might be worth a couple million dollars.
> The heat is waste heat. If it cannot be recovered as a profitable source of energy, [...]
Yes, they should make all deals that make sense for both parties, definitely. But it's only viable in some places some of the time.
https://blog.andymasley.com/p/contra-benn-jordan-data-center...
Water usage is not a cost with this new technology -- that is what the article is about.
Infrasound is terrifically understudied and should not be discussed definitively based on the findings of a highly-biased amateur, but regardless: fan sound is not a cost with this new technology -- that is what the article is about.
Re:pollution, I suppose all buildings are kinda inherently polluting just by existing. So you've got them on that point!
Most importantly, actually: the person above clearly knows about all this, and was just discussing the benefits on their own. I love me some pedantry (really!) but this attempt seems counterproductive, sorry.
hmm? maybe it differs from time to time? on summer, it IS cost (nearby need to run AC)
Aaahh, there’s the catch. “Save resources on cooling by building your data centre somewhere cold, and pollute the surrounding environment by dumping your waste heat wholesale into that!”
Good job Nvidia, I almost thought we had something good there.
"The geography caveat matters. A data center in the Scottish Highlands and one in Phoenix, Arizona, face very different realities. But even in warmer climates, the shift toward 45-degrees-Celsius coolant moves operators significantly closer to that chiller-less ideal — where chillers may turn on just a few days a year when the outside air temperature demands it."
2. Alaska has all of those things. (Though there's plenty of "significant nature" in both Greenland and Alaska.)
https://resources.telegeography.com/telegeography-content-pr...
This shouldn’t be a surprised to the majority of people.
It's not a sleight of hand.
Datacenter locations mainly optimize power, cooling, and fibre.
Cooling is not the same everywhere, even the US has places that are cool enough to not need evaporative cooling water use.
https://yaleclimateconnections.org/2026/06/as-lone-star-tick...
Or malaria.
What a disingenuous comparison, do you hear yourself?
It is impossible for a datacenter to meaningfully heat more than the air in its immediate vicinity.
I mean there are a lot of solutions to making more power or making more efficient chips but to trivialize the impacts of AI as anti Ai propaganda which is how this comment comes of its a bit propaganda minded itself. I think backlash to AI is direct result of strawmaninh Anti AI as having only deranged opinions and anti progress. Where the datacenters are located people suffer because of high velocity don't care about consequences thinking. Every industry has to planing around people or people will do something crazy like burn the centers or try and nationalize data centers. I definitely think that would be stupid but don't underestimate stupid
No free lunch.
Maybe you can't (or don't want to), but people can absolutely fight data center construction. Your lack of imagination doesn't bind others.
Sounds more like a threat than something desired
Ofcourse it is easier when they are all owned by foreign companies lol. If they actually made money and provided jobs (lmao) it might be different.
Not that this is an apples to apples comparison but imagine people saying that about nuclear reactors. Yes, we want them to be as safe as possible and as efficient as possible but that doesn't mean we don't build it. And at the risk of being a NIMBY, AI data centers don't have to be located right next to my house. Unlike regular data centers, these can and should be in the frigid cold. That would probably be for the best anyway. All it needs is redundant fiber network connection, right? Also we should probably require all AI data centers to use renewable energy only. All these are doable but the fight against AI data centers has to be practical and solutions oriented, not fighting it for the sake of fighting it.
It takes one recesssion and all datacenter spending will get rolled back. Yes technology will not go away, AI has been here for over many decades that doesn't mean the current paradigm will continue always and exactly the same forever into the future. If history shows us anything its that paradigms change
I get that they're using liquid coolant at higher than usual temperatures, but why couldn't they do that before? Most of the comparison in the article is for air cooled datacenters but what about other liquid cooled ones?
Surely in all the previous datacenters that have been designed there has been someone doing the math and determining what temperature things need to run at, how much energy it will use, how much heat it all will produce, etc.
edit: just saw this:
>Previous liquid-cooled servers were hybrid: GPUs and CPUs got cold plates, but the rest of the system stayed air-cooled, with finned heat sinks designed to shed heat into moving air. In a fully liquid-cooled server, the cooling for these components needed to be completely redesigned to use liquid.
The rest is marketing: The Cray supercomputer were fluid cooled back in the 1980's, the entire board had an inert liquid flowing across it.
He showed me their Cray, which had its own dedicated computer room, and they set it up with the coolant pump and fountain unit right in the middle in front of a glass wall facing the hallway so everyone could gawk at it.
> The Cray supercomputer were fluid cooled back in the 1980's, the entire
> board had an inert liquid flowing across it.
Does it increases manufacturing and operational cost of such racks?
It seemed like a pretty big deal ~ 2011 when big companies were running their (air cooled) datacenters closer to 95F (35C) vs the traditional 72F (22C). So jumping up a little more is maybe not super exciting, but it's still innovation.
Even air-cooled datacenters work somewhat the same way, but instead of water to chips, it's air. The air goes into hot aisles then exchanges heat with water, after which, see above.
* Other datacenter marketing materials talk about how they have a "closed loop system that uses no water" and they do still use water in the evap towers. I was half expecting this article to be that again, glad it wasn't.
"NVIDIA’s thermal engineering team reworked how those components handle heat, designing cooling loops that simplify how liquid is routed to multiple high-power chips on the board using a single inlet and outlet, resulting in a cleaner tray-level cooling architecture"
The innovation may be in the speed or volume flow of the coolant through different parts of the data centre to regulate the temperature. And of course, redesigning every component to be compatible with this fan-less design.
I think it’s only possibly because NVIDIA is much more vertically integrated than ever before.
So a multitude of communities rebelling and complaints about environmental damage fell on deaf ears but a technical spec might be paid attention to.
What's a favorable climate, apart from, obviously, Greenland? The piece is a little light on details on the correlation between outside temperatures and efficiency & cost. It'd be nice to see even a broad-strokes discussion of that.
https://www.kit.edu/kit/english/pi_2024_038_kit-supercompute...
I assume Germany is the same, many years ago really is different to today.
For e.g you might think of the outskirts of London as fairly moderate, but this week it's been hot enough that supplemental cooling would likely have been needed at points. For a data centre here you'd typically design the cooling system to cope with outdoor temps in excess of 40°C, which is not a conservative number anymore.
Also, while Nvidia might be happy with you supplying water at 45°C I suspect you will get better longevity of the hardware at lower temps like say 35°C. GPUs are expensive, and extending longevity may well be 'worth' a bit more water or energy to you. In practice you are also likely to have air cooled systems that sit 'beside' the AI compute like storage severs, any extra CPU compute and network switches. So you are likely to need a separate room and cooling system for that. Great progress though.
https://www.nasa.gov/centers-and-facilities/ames/doing-more-...
https://www.nas.nasa.gov/assets/nas/pdf/ModularSupercomputin...
Don't the US Military and NASA use metric now?
https://www.sciencedirect.com/science/article/abs/pii/S09213...
https://www.politico.com/news/2026/05/08/georgia-data-center...
and reports of undrinkable water due to the ai center construction?
https://www.bbc.com/news/articles/cy8gy7lv448o
even in the article you linked, it admits AI data centers CAN harm water supplies
"Individual data centers can sometimes stress local water systems in the way other industries do, but when you use AI, you are not contributing to a significant problem for water management compared to most other things you do in your day to day life. "
What is the point youre trying to make?
He also posits in the popular NYT article about a data center sucking up all the water: "But the reason their taps ran dry (which the article itself says) was entirely because of sediment buildup in groundwater from construction. It had nothing to do with the data center’s normal operations (it hadn’t begun operating yet, and doesn’t even draw from local groundwater). The residents were wronged by Meta here and deserve compensation, but this is not an example of a data center’s water demand harming a local population."
And its like he claims "its not ai data centers themselves, its the construction of them" as if its an important distinction that exempts data centers from harm. It's not.
One day, the data center wasnt there, now it is. And the sudden presence of that DC caused the water problems.
But this has nothing to do with data centers, it has to do with construction. I think getting every mad at all new construction in general is a very dangerous path to go down.
A data centre's construction sentiment clogging pipes is fixable; an acute issue. The discourse is that AI causes chronic strain.
But like... the same people will watch 2-3 hours of HD Netflix or post on TikTok right after lodging their complaints on Reddit. The MW/h of AI isn't that big compared to these services. And the carbon sync of digital goods, including data centres, pales in comparison to the supply chain of Blockbuster. Nobody gave af then.
The reason they're cherry picking AI is because we're all sick of seeing tech billionaires thrive when everyone else is pressed by inflation and stagnant job markets.
> What actually happened here was that [...] two water hookups at a data center construction site weren't properly registered or linked to a billable account. When the utility noticed the problem, they sent the data center a retroactive bill for all the water, for $147,474 covering ~29M gallons. The data center paid it.
> Did the data center cause the low water pressure? The article very strongly implies a connection, but doesn't give the reader the right numbers to judge whether this actually happened. To figure this out, we need to know over what time the 30 million gallons was drawn [...] That's about 1% of the county's daily water output. There are a ton of ways a water system like this can experience low water pressure, and a 1% dip just isn't one of them.
> [...] the complaint about low water pressure [which led to discovering the billing problem] was coming from a private well that the data center didn't even draw from
https://www.kpbs.org/news/environment/2026/06/15/imperial-va...
I'll bet anything you'd be the guy that won't drink a glass of the "perfectly safe, completely unaffected" water while making such utterly absurd claims.
The usual way to cool servers is with air and heatsinks attached to the hot hardware, similar to how your desktop computer or laptop works. As the hardware gets denser and more powerful, you need bigger and bigger heatsinks and cooler air blown over them. At some point you can't make the heatsinks bigger because of space constraints, and you can't blow the air faster (because of noise and efficiency), so you need cooler air. That's when you start running chillers that evaporate water to cool your intake air. This is the huge water consumption that we'd like to avoid.
The next step is, obviously, liquid cooling. Again, this is similar to your fancy gaming desktop. You can dump a lot of heat to a liquid medium through a small heat exchanger inside, where you're space constrained, and you can run the liquid through a gigantic heat exchanger outside, despite the temp delta between your coolant and outside air being pretty small.
This article is about a system that's FULLY liquid cooled — CPUs, GPUs, memory, networking, the whole thing. That's the actual cool part (pun unintended). On top of that, their solution is optimized to be able to run the coolant quite warm — this obviously limits the heat flux at the hardware side, but it allows you to run the outside heat exchangers "dry", i.e. without wasting any water for its latent heat.
You realize how that vague and open ended clause completely undermines any point you are trying to make, right?
Some of the folks looking to build datacenters to cash in on the AI craze aren't as bright as that.
Our data center is mostly air-cooled, with 100F hot aisles and chilled water going to heat exchangers in every third rack position, although some of our newer GPU racks use chilled water directly. For most of the year that means we don't need chillers - the warm water return is sufficiently hotter than outdoor temp that heat "flows downhill" with a bit of pumping.
But we use an evaporative cooler to carry that heat away outside, because it's more efficient. If you look up the heat of vaporization of water, that means we evaporate about 10,000 gallons of water per day per megawatt of power dissipated. We're on canal off a large river, and we don't dissipate all that many megawatts, so I believe the water use isn't significant.
Using air as the working fluid to draw heat from your machines has a limitation - humans have to breathe that air whenever they work on the equipment, so the temperature is limited. Once the exterior temperature nears your hot aisle air temp, you either need active A/C to create a heat source hotter than ambient, or an evaporative cooler to lower the "effective" ambient temp to the dew point.
Liquid cooling lets you run your working fluid a lot hotter without killing folks like me who go into the data center, although honestly 45C sounds like an incremental improvement over the 100F places like ours are already running. (although to be fair, the warm water return from the heat exchangers is no doubt somewhat less than 100F) It also lets you run your "cold" side a lot hotter - if you "chill" your water down to 100F (38C) on a hot day, it's still cold enough to carry away a lot of heat at 45C.
(I'm skipping over the fact that there are multiple heat exchanger loops involved - e.g. any system with an evaporative cooler needs a heat exchanger to keep leaves and bird shit outside the building where it belongs)
I disagree if there is an assumption or premise that all datacenters use water for cooling. Datacenters are not built or operated the same, nor have a requirement for water cooling, or evaporative cooling.
Water use is partially a myth for sure, because it's not a certainty or requirement in every datacenter, ever.
If water is used for evaporative cooling, it's used in places that the cooling can't work above a certain temperature. Say, for example the US where it gets over 120 degrees farenheit.
The inability for cooling equipment to perform when it's extremely hot might be supplemented my multiple technologies cooling together, including evaporative cooling. So evaporative cooling might be common in the US or other regions like it.
There are many datacenters that don't need to consume water at all for cooling, except for a water cooler to drink from, and sprinkler systems for fire suppression. They are engineered and cooled just fine.
The demand for datacenters and money in this space will likely solve the water usage, if for no other reason, than to increase flexibility of locations.
If anyone is curious a bit more about how:
Datacenters locations optimize 3 main factors.. Power, Cooling, and Fibre.
We know the climate of location isn't the same everywhere in the world where datacenters are located.
One thing that can get missed among the water hyperventilating crowd is that datacenters are designed for the need of the location's climate, and what it takes to cool. Too many folks spreading the information are usually non-technical, and get their education about datacenters from social media.
There are many places in the world, including in Canada, where cooling is built in 6 months of the year to outdoor air. The temperature isn't high enough in the summer to warrant or require evaporative cooling. They can just vent outside, or capture the heat for other purposes.
In other areas still, proven building approaches such as countersinking the datacenter for some additional geothermal benefit, local power generation (Solar, Natural gas) and a unique path to the same datacenter is possible.
Of course, if foreign companies are wanting to build in a different country, they might by default build how they know for their region and assume it's ok.
There is a massive build out of datacenters occurring in Canada right now, they will be built according to how buildings are built in that climate, not importing cooling technology from regions where it's too hot, or evaporative cooling is the only, or most available option. In temperate climates, considerations have to be made for keeping keep buildings warm (not just cool) year round.
The nice thing is this is not hypothetical.
There is of course, equipment that can have liquid cooling, or we see those fun videos of equipment submerged in liquid operating. These seem to not be the common.
My point is not that evaporative cooling is necessary - it's that it's efficient, and therefore common. I found some 10-year-old info from our datacenter, and it looks like mid-summer months add an extra 20% in power consumption for cooling, in large part due to running the chillers, which are industrial-scale active A/C units to raise the exhaust heat temperature. Average outdoor temp (24-hr average) during those months was 70-75F.
It's a pretty small datacenter, so getting rid of the evaporative cooler would probably "only" cost us a couple of megawatts or so in additional power consumption. (and maybe 15 million pounds of carbon emissions - our power is hydro, but electricity is fungible, so every kWh we use probably results in an additional kWh of fossil fuel generation somewhere in the grid)
Or are they for some unfathomable reason using evaporative cooling in data centers?
East of the 100 degree W line of longitude, there is more than enough water to use evaporative cooling if needed.
currently.
and also more energy efficient, because evaporating water away takes a lot of energy with it. you have to raise radiators to a higher temperature to keep up with that, or have much more surface area.
The loss of material must be included
If water is evaporated or spent out of the system.. it is not more efficient
You can certainly argue DCs should pay more for water than other uses, but who gets to decide what is a good vs bad use of water? Pricing in externalities is tricky, and water usage rights are especially complicated. I don't know what a good&fair solution is.
https://www.sciencedirect.com/science/article/abs/pii/S13594...
A closed circuit cooling tower still has water spraying onto the closed loop process water heat exchanger coil and mixing with atmospheric air to evaporate and cool the process water indirectly instead of evaporating and recirculating the process water that doesn’t evaporate directly like in an open-loop cooling tower.
I suppose you could condense the evaporated water somehow by using a chilled umbrella or some other ridiculous contraption above the cooling tower, but why would you do that?
FWIW I sell and run commercial electrical work, primarily to mechanical contractors who are installing boilers, chillers, cooling towers, and pumps. I spend my professional life immersed in this type of equipment.
1gw of power converts approx 400 liters of cold water into steam _per second_.
Historically, cooling alone has accounted for up to 40% of a data center’s electricity consumption
They’ve made this claim numerous times in the article and I really don’t understand it. The building has tons of water being recirculated through it. That water came from somewhere in the surrounding natural world. How is that 0 water consumption?
But it's still misleading. The major source of water use in datacenters, by _far_, is the water used in power generation. This improves PUE, which reduces power draw, but the savings are almost certainly under about 20% given that many modern datacenters already operate at a PUE of under 1.2. So if you're running on coal or gas, you're still consuming quite a bit of water indirectly.
Now that said again - the water consumption part of this equation is generally overhyped. The power draw is the problem, as are the really bad temporary hacks to the power problem (e.g., what x.ai is doing with "temporary" gas turbines).
A PUE of 1.2 means that for every watt of power you're using in a server, you're using 1.2 watts drawn from the grid (approximately). Those extra 0.2 watts are being spent on cooling and other things.
Imagine a residential building that reclaims and reprocesses and purifies 100% of all the water it uses. This would be dramatically more difficult and better than the status quo, and would be called 'net zero' by any sensible accounting method.
---
Obviously, evaporative cooling is net-zero water use when accounted across the entire globe (the water falls as rain, somewhere, eventually), but it is net-negative for a local community.
But up north there's at least one city (Kajaani) on data center heat; details in another comment.
That doesn't mean a datacenter doesn't have a very high _initial_ need for water, or to replace _some_ amount of leakage, replacement etc (agreed it will be way lower than say a swamp cooler). For example, they could be using millions of gallons as a sort of "ballast" to keep the water temperature very stable in the short run.
This whole projects depends on the whole stack of components to be able to tolerate 45c from memory, drives, the lighting of the building, the humans ...
There are some systems that pipe refrigerant around the building, but they’re relatively uncommon (VRF or variable refrigerant flow if you want more details).
Glycol and water is cheaper than refrigerant so there’s usually a chilled water loop that passes thru a heat exchanger that interfaces with a chiller (vapor compression refrigeration) to reject the heat from the chilled water loop.
This eliminates the need for evaporative cooling towers.
It strikes me that building everything around room temperature or slightly chilled air is a strange choice. This is already 290K-300K or so, and now this is suggesting that things run fine at 320K or 330K?
I've wondered why we couldn't just design everything to operate around 200°C and just use free-cooling by pumping ambient air through. Why don't data centres look more like chicken barns? Do things melt? Are there more errors of some other type at high temperatures?
Energy being energy means that high enough ambient heat can kick electrons into the higher orbitals because the band gap is so small. This also happens at normal ambient temperatures, but those electrons don’t make it very far and there aren’t that many. At 200c a closed gate is not stopping enough electrons from moving through it anymore.
At least this is the slightly hand waved technical explanation. Project in Flight on Youtube has an excellent video on how semiconductors work.
It would require entirely different chips with entirely different manufacturing processes.
Does that mean the whole server board is running in liquid? If true, how do they do maintenance? Replacing parts must be extremely difficult.
But it did not sound like they were describing a Cray 2 style liquid immersion cooling system.
On the other hand: the heat has to go somewhere. So… where? Datacenters already create a warm microclimate in their vicinity, is that getting even worse?
It’s kind of like how brine from desalination is not a global problem for the oceans at all — all that matters is diluting it enough that it doesn’t poison the local ecosystem.
It's not clear to me what changes are happening here. The siblings to your post seem to be indicating an overall improvement.
If you manage to use the waste heat to avoid generating heat somewhere else (that the article calls heat recovery) then there’s a further reduction in total heat output.
More on it at [0], but it doesn't take anything beyond a basic energy calculation to know that 1GW of energy is not going to have a significant effect "6.2 miles away".
[0] https://andymasley.com/writing/data-centers-heat-exhaust-is-...
Look at the author of the blog.
https://blogs.nvidia.com/blog/author/joshparker/
Is this NVIDIA lawyer trying to influence the public perception of datacenter for his company?
Does anyone have any sources for this sort of data?
What you could do is to use a heat pump to spend a little electricity and turn the waste heat to hot water usable for heating (winter).
I'm ignorant to the specifics of how this might work (and whether it'd even be feasible at a cost/logistics level), but it feels far less invasive and far more sustainable as an armchair thought. Curious if anyone has worked on something like this and what sort of gotchas you've found.
Leslie Groves, Oppenheimer
Nvidia has so much money and they can’t afford to pay a human for a day of their time to write a blog post?
> Nvidia has so much money and they can’t afford to pay a human for a day of their time to write a blog post?
The shareholders desperately need that money.
We are all fucked.
And it’s sad because Jensen seems like one of the rare good CEOs when I listen to him speak.
But even Dario says he doesn’t let Claude actually write his blog.
Have we been listening to the same person speak for the last few years? Jensen rarely even sounds sane anymore.
The same as technical docs for any codebase, humans will not read them anymore, only AIs which then translate it to human on-demand, it's already happening, I've worked recently with many new frameworks/codebases without even opening the doc (not even the Github page) and solely asking the agent to gather info for me about it.
PS: The reason I feel it will be this way is that it will allow to legitimatize mass data collection indirectly, instead of doing telemetry on page and software level, we will just send all the content automatically to some inference providers (probably provided for free by Google, MS and so-on)
Watch for cases where content has been through two layers of LLMs. It's not good.
I do think ground based centers will have better economics when they can be built though, and this addresses noise and water complaints which are the big 2 regional complaints.
It seems like lots of bottlenecks are getting solved quickly, except for maybe memory.
https://spacenews.com/china-backs-orbital-data-center-startu...
https://www.reuters.com/science/china-vows-develop-space-tou...
Plans for space data centers should be seen with skepticism. However when they are backed by different parties who have stakes in the game, that's more credible. More than HN crowd for sure.
1. Space is terrible for heat regulation. It's a perfect insulator for everything except radiative cooling, which is the least efficient. Hot things stay hot.
2. Space is full of radiation. Everything has to be radiation hardened, which makes it heavier, more expensive, and, yes, more difficult to cool.
3. Space is far away. Well, farther than a data center on Earth can be. I know China hasn't solved the speed of light.
We put up with it with satellites because it still has some advantages over trying to run cables literally everywhere, but we do, in fact, still use cables laid on the bottoms of the oceans.
So, is physics wrong, or is a country known for making dumb decisions some times making a dumb decision?
To answer your first question, yes. Even if they were US companies. Remember when each Big Tech player was claiming to build their own blockchain, except centralized which was a contradiction.
I do really think that if large numbers of jobs are indeed going to be displaced by AI, movements will pop up of people attacking datacenters (and honestly I wouldn't blame them even though it won't really accomplish anything). Having them in space keeps them out of reach of anyone but state actors.
To be precise, heat rejection via radiative cooling scales with the fourth power of the temperature (in K) the radiator operates at, all else constant.
If we pick an extremely fast orbit, then relativity means the hardware will age out (slightly) slower, so I'm sure that'll help with the maintenance issue.
It's the wrong way around though. Ideally we want to speed up our current compute ability not slow it down; if it experiences more time than we do then it can do more. Relative-MHz means my slower hardware becomes tangibly fast again.
General Relativity says mass warps space time, so we need to get these datacentres out of the Earth's gravity well. And the Sun's, and the Milky Way's; out into the deepest void of intergalactic space. The good news is that a maintenance callout is still quicker than some of the earth based DC's I've had gear in, but the bad news is that it doesn't get us much of anything at all.
Special Relativity lets us abuse time with speed (something I discovered as a teenager). Going faster than Earth means we experience less time, so we just need to try and slow down comparative to our home base. The earth is orbiting the Sun at ~30km/s, the solar system is orbiting the centre of the Milky Way at ~230km/s and our local group of galaxies is moving relative to the Cosmic Microwave Background at ~600km/s. We can easily get our DataSpaceCentre up to 1,000km/s or more, so we just need to point it relative to all that movement we mentioned above making stationary relative to the universe. It's completely doable, but (as well as far more variable response times to callouts) only gets us an extra second of compute over a human lifetime.
Fundamentally, we're attacking this problem in the wrong direction. Earth's gravity is comparatively minor, and our piddly ~600km/s relative movement is a tiny fraction of the speed of light. We should be filling The Earth with compute, and then decamping humanity into space and travelling at relativistic speeds. Or put the compute in space and move the Earth into the event horizon of a black hole. You can't do the inverse of Interstellar keeping Earth where it is, the maths isn't in our favour. If everyone lived on (a less moist) Miller's Planet, we'd get 7 years of compute every hour. It puts Moore's Law to shame; the relative MHz are obscene.
There's the obvious problem of communications. I'm led to believe there's issues with radio and light, so this probably isn't a job for fibre. Veritasium seemed to imply a battery, switch, lightbulb and a wire stretching around the globe would light instantaneously, so I'm sure we can come up with a new copper Ethernet standard for low latency over solar distances.
Invest early, we're going straight past the moon!
The only reason is to remove them from local/regional, and potentially national/international jurisdictions.
Oh, and of course, the other reason is to give SpaceX to have a reason for xAI to be part of its structure.
Starlink makes sense, LEO comms using a mesh makes sense.
LEO/geosync satellite data processing doesn't.