Like, would a skyscraper-style datacenter be practical? Or is just a matter of big, flat buildings being cheaper?
Most obvious reason that comes to mind is cooling. All the HVAC in the world isnt going to be able to stop 50 floors of server rack heat from spilling onto the 51st floor, etc.
This is also a nightmare for fire suppression systems.
Weight. If you load a 42u rack up with 30lb servers you’re at 1280lbs spread out over about 4sqft, which is over the floor loading limit for most buildings. It’s much cheaper to support the weight in a wider building compared to a taller one.
That being said I’ve been in many data centers in the middle of a giant office towers, but they have lower weight limits generally.
This is armchair nonsense.
Slab reinforcement in skyscrapers can literally hold fucking skyscrapers.
If you want to call this a limit, the limit is the expense. Not the weight. That’s absurd. We’re so far past that.
There are a few skyscraper server farms, but they are usually in major metropolitan areas like Manhattan. But, as you touched on, it is usually cheaper to build out instead of up.
Electricity has a hard time flowing up and requires a special pumping system.
Which begs the question why not magnets at the top of the building to help pull the electricity up?
Because nobody knows how magnets work.
Solution: instruct the buildings upside down, so the foundations are up in the air and the roofs are underground. That way, the electricity will flow down instead of up.
But then the roof has to support the entire weight of planet Earth on top of it, which is a much harder engineering challenge than pumping the electricity in the first place.
No, it’s not. It’s all empty space under the foundation. There’s nothing to create crushing force against the building.
You are failing to account for the weight of the atmosphere on the foundation
The atmosphere is just air. Air doesn’t have mass or weight, that’s why it floats.
Because the electricity pulls the magnets down in the same measure, so they meet in the middle. Newton’s 2nd law or something.
You joke, but heat does rise and a tall building would need to make extra concessions for cooling concerns, while also dealing with the issues if weight. Large racks of servers are actually quite heavy, which is why many datacenters in i.e Toronto were built in an ex parking garage
Heat is no issue of relevance for the question. The rising effect it negligible compared to what has to be transported anyway. I also can’t imagine racks being heavier than eg books in s library.
Unironically, I’ve had people telling me they save electric energy by inserting the angled Schuko plugs of their electric devices ‘upwards’.
I used to have a rack in a colo in the middle of Portland Oregon. The 3rd floor was the colo and they also had telecom equipment on the roof, but the rest of the building was normal commercial real-estate.
They spent so much reinforcing the floor, sound proofing the floor and ceiling, building a giant door in the wall (so they could crane in equipment), and helping pay for a new local substation that when P.a.a.S. services (AWS, Azure, etc) started they quickly started to struggle to keep the bills paid.
Soundproofing wouldn’t be needed if the whole building were servers, but the rest would be.
This. It’s the weight.
Tall data centers do exist in cities where land is expensive. It’s the same reason everyone builds up in cities. Where land is cheap and available it’s usually easier and less expensive to build things low and wide.
Tall data centers do exist in cities where land is expensive.
Probably a bit of “hiding in plain sight” that way, too. There are a few big datacenters relatively near me, and they’re massive compounds in the middle of even more massive corn fields. Kind of stick out like a sore thumb when you’re driving by.
Cost. It’s slightly cheaper. And I do mean slightly.
Depends on the price of the land.
Yeah that’s the only difference. Usually land is cheaper. But building high in the middle of nowhere is only like 5% more expensive.
That’s complete bullshit. Building one floor on cheap land is way, way cheaper than building high on expensive land.
source: Wyoming.
and of course, building one floor of data center is exponentially cheaper than building two floors.
It’s also not the core purpose of the data center. The building is a shell protecting racks of servers, the core purpose of a data center. Why spend anymore than absolutely necessary when it would not increase the output of that core purpose.
In the past, we did have a need for purpose-built skyscrapers meant to house dense racks of electronic machines, but it wasn’t for data centers. No, it was for telephone equipment. See the AT&T Long Lines building in NYC, a windowless monolith of a structure in Lower Manhattan. It stands at 170 meters (550 ft).
This NYC example shows that it’s entirely possible for telephone equipment to build up, and was very necessary considering the cost of real estate in that city. But if we look at the difference between a telephone exchange and a data center, we quickly realize why the latter can’t practically achieve skyscraper heights.
Data centers consume enormous amounts of electric power, and this produces a near-equivalent amount of heat. The chiller units for a data center are themselves estimated to consume something around a quarter of the site’s power consumption, to dissipate the heat energy of the computing equipment. For a data center that’s a few stories tall, the heat density per land area is enough that a roof-top chiller can cool it. But if the data center grows taller, it has a lower ratio of rooftop to interior volume.
This is not unlike the ratio of surface area to interior volume, which is a limiting factor for how large (or small) animals can be, before they overheat themselves. So even if we could mount chiller units up the sides of a building – which we can’t, because heat from the lower unit would affect an upper unit – we still have this problem of too much heat in a limited land area.
Probably a combination of it being cheaper (as the others said) and data centers not really having to be near right next to cities. It’s easier to plop them in the middle of nowhere.
Taller is more expensive. That’s all.
If you have the space to go wider, that’s what everyone does.The taller it goes the more they’ll have to value human life.
Well, data centers require a lot of water for cooling. It’s harder to pump water up high.
It’s harder to do anything up high. Construction, climate control, transportation (moving something from one end to another), managing fires, additions, evacuation of personnel in an emergency, etc. etc., all get much harder when you build up up versus across.
Guess it depends on the height, but yeah. Otherwise, we manage to pump a town’s worth of water to the top of a tower well enough. From there, gravity can do the rest.
But there’s probably a point where cost for that vs height becomes prohibitive.
There’s a slow, steady source of water being pumped into those towers, so the water level fluctuates through the day. They use gravity to get high pressure from your tap, but I’m the world of liquid cooling, you gotta go fast!
Cheaper and easier, perhaps? If the costs of engineering a tower is more than just buying more land, then why build taller?
It could also be local ordinances. Maybe they can’t get permission to build higher than 2 or 3 stories. I know that’s common for residential areas; it might be a thing in commercial or industrial areas too.
If the costs of engineering a tower is more than just buying more land, then why build taller?
Figured it’d be something like that. Explains why they get built out in the middle of nowhere since land is cheap.
