tax, money, etc, supplying electricity to average joe doesn't pay for powerplants, big corp with billions of dollars can
also probably political donations etc, majority of infrastructure are not built for average people actually, not just modern llm problems

Overloading the power grid to compete with China's patent quantum spending and starving people for energy is better than China using it for it's expansion?

>>16951489
China is an irrelevant paper tiger based on tik tok horse shit

>>16951403
Data centers unlike AC's consume power continuously.
>>>/pol/

>>16951403
Even if everyone was generating shit non-stop all day with AI it would be using like 20x less energy than all the air conditioners. It takes a ton of energy to cool a whole house.

>>16951403
If you can compel them to have skin in the game, you can force them to fund infrastructure upgrades after they have already made a significant structural investment.

>>>/sci/ai+containment+thread

>>16951523
You dont know what a paper tiger is.

>>16951559
That makes it worse though why would the grid be able to hand more year round consumption if it can't even handle summer consumption?

>>16951596
Are you blind, did you skip OP's image? Most of the energy to run AI servers is to power their AC units. Why would it take 20x less energy to cool a whole data center than a whole house when data centers are well over 20x the size of a house?

>>16951688
It'a harder to accommodate spikes than continuous consumption. Each days spikes could vary from the next and then there's the fact that thousands of humans are the ones who blare up the A/C means that even if you do have a somewhat modern electrical system and grid on hand if people decide to turn their homes into Hong Kong malls by pushing the AC far beyond effiency specs you can still get fucked.

>>16951559
>>16951671
This is one of the only science related threads on here at the moment, its fine

It probably has something to do with corruption. BUT it is worth considering that AC units tend to all turn on at once so there is very high demand at some parts of day with none at others, while data centers have a nice steady draw that is much nicer for the power utility to deal with.

>>16951698
>It'a harder to accommodate spikes than continuous consumption.
So data centers don't also spike when it gets hotter outside and they don't spike during more active times like when something "breaks the internet"?
>by pushing the AC
So summer spikes in energy demand are primarily caused by people's choices rather than by the ambient temperature spiking dramatically?

>>16951711
What exactly makes an AC unit behave differently when mounted inside a home vs mounted in a data center? Why don't they just make home ACs like data center ACs so the power draw is similarly steady?

>>16951688
No it doesn't. When a data center is consuming electricity continuously a power provider will just build a new power plant to generate that power. When AC is used for 30 minutes during the hottest day of the year there's no way for power providers to accommodate for that except by having tons of power plants doing nothing for most of the year.
>>>/pol/

>>16951720
Data center AC units run continuously. If you run AC on the same blast in winter as you do in summer then your AC also doesn't harm the grid.
>>>/pol/

>>16951724
>Durr no it doesn't make it worse, it just makes demand so bad an entire power plant has to be built.

>When AC is used for 30 minutes during the hottest day of the year
So data centers don't get hot when it gets hot out like a house does, so they don't also have demand spikes when the outside temperature spikes?

>Data center AC units run continuously. If you run AC on the same blast in winter as you do in summer then your AC also doesn't harm the grid.
And they don't have to consume more energy when they have to cool down more because the outside temperature is spiking or they are fine letting their servers all overheat in the summer to maintain continuous power draw?

Are you trying to redirect to another board because you clearly don't understand basic thermodynamics and don't want to be held to a scientific standard?

>>16951731
this isn't thermodynamics nigga just look at a demand graph, this is electricity industry 101

>>16951731
A data center can't turn off the AC during the winter, the chips would melt instantly. The heat is coming from inside the house. The exit temperature hardly matters for the electricity consumption of a data center AC units.

>>16951734
Sure, the energy requirements of temperature management has nothing to do with thermodynamics, you figured out the secret to air conditioning, just electrocute the air with 1.01 gigawatts.

>>16951736
>A data center can't turn off the AC during the winter,
But it can use more energy to run the AC to maintain the constant temperature during the hotter summers, dipshit.

>The exit temperature hardly matters for the electricity consumption of a data center AC units.
So why does the outside temperature affect houses so drastically, but not data centers?

>>16951740
The AC runs continuously at the same blast because otherwise the chips will melt.
>So why does the outside temperature affect houses so drastically, but not data centers?
You don't have a gigawatt of power production inside of your house. The heat is coming in from the outside.

>>16951741
>The AC runs continuously at the same blast because otherwise the chips will melt.
If that blast cools them correctly in the winter, why wouldn't there need to be more in the summer when it is hotter and easier to make the chips melt?
>You don't have a gigawatt of power production inside of your house.
So they just turn off half the servers to produce less heat when there is more heat coming in from outside, so they can use "the same blast" to cool the chips so they don't melt?

>>16951741
>You don't have a gigawatt of power production inside of your house
How much do I have in my house and why doesn't it follow the same thermodynamic rules as the heat inside a data center and how is a gigawatt not a measure of heat over time?

>>16951686
>Paper tiger" is an idiom for someone or something that appears powerful but is actually weak or harmless
Seems to describe China very well

>>16951744
You are genuinely retarded.
The temperature range to keep a data center operational renders the outside air a moot point. The coolers are running constantly, at the same rate, without any regard for summer or winter temperatures.
This uses more energy overall. But consistent high draw is preferable to inconsistent moderate draw.

>>16951403
Data centers are evil.

>>16951747
When cooling a building two of the major factors are how much heat is entering the building from the outside, and how much heat is being produced by things inside of the building (either electronic devices being used, producing waste heat, combustion devices like gas stoves, people or pets).

With a data center, the amount of heat being produced by the computers is much, much larger than the amount of heat entering from the outside. So the difference between cooling requirements on a cold day or a hot day is rounding error; if you are already removing a steady gigawatt worth of heat, having to push out a few extra megawatts of heat coming in on a hot day is nothing.

With a personal residence, the amount of heat being produced inside by various devices and people is pretty trivial. On average maybe a kilowatt or so? On a hot day there will be several kilowatts of heat seeping into the house, much more than is being produced inside.

Basically a hot day increases the power use of a data center by a few percent at most. A hot day at a house can easily increase the power use by 400%-500%.

>>16952105
They aren't. They've existed for decades anon.

>>16951724
One must mention it's far easier for a energy company to strike a set contract with specific rates with a data center company than doing the same with thousands or millions of customers.

>>16951717
>So summer spikes in energy demand are primarily caused by people's choices rather than by the ambient temperature spiking dramatically?
You don't really need that much cooling to have a cool home. Going beyond it leads to dry uncomfortable air and going beyond that leads to wnergy inefficiency because the AC unit has to expend more energy to cool the air even further.

maybe it also has to do with wires. like, they will make special high-powered wires from the power plant to the datacenter, but in residential areas they can't go beyond what was built originally and that has become insufficient with increased usage

All you retards are arguing cost effectiveness when the original point was about overloading the grid. If AC units on the average home is too much to handle, then proposing data centers ontop of this makes literally no sense.

>>16952629
republican states often don't have that much sense to begin with so it's right on character for them

>>16952639
I live in Michigan where this is happening and the governor is a Democrat.

The same Democrat that had the harshest COVID lockdowns in the nation and still finished with a worse death rate than Florida. Fun fact.

>>16952201
This. I lived with a buddy in college who had a single server in his room and the thing was like a fucking space heater. Hed have to open his windows in winter because his room was baking.

A data center has to have massive cooling capabilities.

>>16952629
See >>16952218
Ever been to a gym and noticed how many of them have cold by super dry air? That's basically pushing for more coolness far past the point of diminishing returns. Even if we did roll out more effiencient AC home cooling as long as people and retail/offices keep spiking the system then energy infrastructure problems will continue to occur. Yeah we could build specific facilties to handle AC spikes but that too requires more energy infrastructure money.

>>16952629
>then proposing data centers ontop of this makes literally no sense.
Data centers have continuous load, they don't pose a problem for grid stability because of that.

>>16952654
>harshest COVID lockdowns in the nation and still finished with a worse death rate than Florida.
Is that supposed to be bad?

>>16954687
If you placed lockdowns over COVID and ended up with a worse death rate than a state with 0 restrictions then what would you think? Stupid fuck

>>16951691
> Why would it take 20x less energy to cool a whole data center than a whole house
There's more than one house with AC in a city hosting datacenters, moron.

>>16952629
So ban or limit residential AC usage?

>>16956340
Or you could kill yourself to reduce carbon?

File: china reports fusion results.webm (2.7 MB)

2.7 MB WEBM

>>16951489

>>16953041
>have continuous load,
So i should run my ac 24/7 and this will be better for our grid. Good to know.

>>16956365
My city doesn't have any power issues.

>>16957248
Well, if you and everyone else ran their AC full blast all the time, and were willing to pay for that, then yes, that would be easier for the grid to manage. New power plants and transmission lines would have to be built, but that is manageable when people are paying. The issue is that people want extra peak capacity but don't want to pay for more than average capacity prices.