Click here: to donate by Credit Card
Or here: to donate by PayPal
Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794
Thank you very much and God bless you.
Posted on 07/24/2026 8:17:43 AM PDT by econjack
Perhaps the biggest obstacle to data centers seems to be water use. I just saw this:
"One of the most common questions we get about AI data centers is how much water do they use? Fair question. Older designs cooled servers by evaporating water and people naturally assume ours do too. They don't.
IREN's sites are different. Direct-to-chip liquid cooling in a closed loop. Same idea as a car radiator or a fridge. Fill it once and the water just goes round and round.
Over the life of a 200MW site, the initial fill plus every top-up averages out to about as much water each year as 3 households.
Not a town. Not a suburb. Not a street... 3 houses.
Water matters in the regions we build in. Which is precisely why our cooling barely uses any."
It's an Australian company, and I haven't found much info on it.
|
|
Click here: to donate by Credit Card Or here: to donate by PayPal Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794 Thank you very much and God bless you. |
IREN Ltd (formerly Iris Energy Ltd) is an Australian company headquartered in Sydney, New South Wales. IREN operates data centers that are supplied entirely with renewable energy. Its infrastructure is optimized for both cryptocurrency mining and artificial intelligence (AI) cloud services. The company’s activities include the operation of its own Bitcoin proof-of-work mining facilities, the development of GPU cloud and colocation services, and customized data center projects for large customers. IREN has been listed on the Nasdaq since its initial public offering in November 2021.[2]More AI that isn't Artificial Intelligence.
“Same idea as a car radiator or a fridge. Fill it once and the water just goes round and round.”
~~~
You still have to cool the water through.
So to use the car radiator example, that water that circulates through your engine is cooled by the air flow through the radiator when the car is moving or a fan when it’s not.
My gaming computer loops a liquid over my CPU, but that doesn’t mean my computer is fan-less. It just doens’t have a heat-sink with a fan directly on the CPU itself. The liquid is circulate to a separate heat-sink(s) that has fans on it (them).
The kinds of data centers the use a lot of water are the type that don’t circulate them but pull them from a water supply much like they do with cooling for nuclear reactor sites. If they can’t pull it (and filter it) from a natural water supply, then they have to use municipal water supply, which is costly if they are just dumping the water back into the environment or drain/sewer system and always buying more.
Sounds like they use some kind of chillwater system. The problem is those systems require a refrigerant plant, which means they will need yet more power. The waste heat isn’t going to remove itself.
They use far less water, and far more power as a result.
They are a dry hotspot.
Except they contradict that in the next sentence.
Over the life of a 200MW site, the initial fill plus every top-up averages out to about as much water each year as 3 households.
If you have a closed loop no refill would be necessary.
So the loop is only sort of closed but not really.
Now, the big question, where does your lost water with all the additives go?
Because if it is going up into the atmosphere there might be another tiny problem.
Nothing is perfectly closed. Back when dtr worked in a computer shop, she would get big desktops with closed loop water cooling, it would run out after about three years, so the smart thing was to bring it to the shop annually and let them break the seal, flush and refill the water, and then reseal.
Now, coincidentally enough, dtr now is employed by a company that installs networking at data center sites, so she works alongside IREN people and sees the similarity to the water cooling computers, only on a more massive scale.
Tiny amounts of fluid can escape over time through pipe joints, valve seals, and pump packing glands.
When opening the pipes to repair or replace servers, chillers, or internal components requires draining parts of the system.
sometimes you drain from the bottom of the cooler a small ammount of the circulating liquid to remove built-up minerals, impurities, or treatment chemicals that concentrate during recycling, but the total losses per year are in the hundreds of gallons not thousands.
it is not a big deal.
Msft has several datacenters that use ambient air 99% of the time, they only have to use their chiller 3-5 days per year, I have taken tours.
It depends. If their heat exchangers are liquid/air that’s one thing. If their exchangers are liquid/liquid that’s a whole different thing. They leave that out of the details.
There is always some loss from refreshing coolant and doing maintenance. The same happens in cars. You do not keep the water in the radiator for the life of the car.
Thanks for putting some truth nuggets out there. I know the locals are hysteria driven, just like global warming, and probably from the same folks - China and our biggest competitors are paying for this negative propaganda because they don’t want the US to control it.
Yes. That was my point.
If you tell people it is a closed loop and after they find out it isn't really they are going to believe you lied to them.
The people you are talking to are not tech experts, they will believe that what you say is what you mean. Tell them that it is a closed loop when it is not and someone like me can come along and point out that it is not a closed loop they are not going to believe a thing you say from that point forward.
Your car's radiator is a closed loop, but you still need to add coolant to it occasionally.
So the loop is only sort of closed but not really.
Now, the big question, where does your lost water with all the additives go?
Where does the lost coolant from your car radiator with all the additives go?
I think data center water usage is closed geothermal loop. It’s for most part a one time use.
It is a mostly contained system but not a closed one.
To continue with the car anology, a closed system like the sealed transmission system, where the fluid does break down and so needs a flush and replace but does not need topping off is what most people will think of.
When you are trying to persuade people you need to be precise in your language so they do not think you are trying to pull a fast one.
They have seen way to much of that already.
No. It is not a closed loop. It is a contained loop.
Where does the lost coolant from your car radiator with all the additives go?
Up into the atmosphere. Or on the ground if you have a leak. But because it is a small amount spread over a large area it is generally not a problem.
The solution for pollution is dilution.
This would not be a small amount spread over a large area but a large amount spread over a small area.
This could indeed become a problem.
BS, HOW MUCH WATER IS REQUIRED TO FILL THESE SO CALLED RADIATORS?
And they do make heat.
That is not always appreciated.
When I was in law school, I worked for a utility doing research, and we were getting the permitting in place for the construction of peaking plants - which were effectively a jet engine driving a generator, housed in a metal-sided building that looked like a small pole barn. They were expected to run several hours a year on average. Created a fair amount of opposition when proposed, but were approved and constructed, and probably are the building that most people do not notice unless they are looking at aerial images. It has a driveway and parking for perhaps 4 vehicles. About once every other month, someone comes out to inspect, but no one “works” there. Once a year, typically in August during a heat wave, the utility might face a peaking event and fire it up for 30 to 90 minutes. Otherwise, it is like a cemetery, very quiet. And yet there was a lot of opposition to that generation asset being built.
I think there is a lot of similar irrational opposition to the emergence of data centers. For some reason, even though everyone pays for utility services, essentially no one outside of the industry understands how rates are designed and approved by regulators (with public input). The costs of supplying data centers will be borne by data centers, not by residential users. That is a normal feature of utility rate regulation in every state. It’s been true for nearly the last 100 years of utility regulation in the United States. When you see someone rant about data centers driving up the cost of electricity for residential ratepayers, that is ignorance on display.
Data centers are unusual developments. They require a lot of labor during construction; they add significant value to the tax base; they have very few regular employees once operational. They may not be pretty, but they are easy to hide behind landscaping.
I’m not sure what the controversy is about, but to the extent that some states reject data centers, those that accept them, like TX and FL, will gain a lot.
Exactly! They’re saving water and using more energy.
Proving again there is no free lunch.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.