Investment

So there's no stuff to heat up and move away from your data centre carrying the heat.
You can use radiative emission. That sounds good, because , if you remember skule fiziks, it's proportional to the temperature difference to the power 4. Do the names Stefan & Boltzmann ring a bell?
Your chips run at say 50C which is 323K
Then you plug in the constant, which is 5.670367 × 10⁻⁸ W·m⁻²·K⁻⁴
Spot the pesky -8 in there.
It comes out that you need about 1 order of magnitude less for the size of radiator to cool the thing, than you needed to get the energy.
And they're really heavy, compared with solar collectors. You need a couple of thousand square kilometers of the things, all got there by rocket.

Oh, and the data centre itself would be pretty big and heavy. Maybe 100 square kilometers, but they're definitely three dimensional.

Then there's the maintenance problem
and bits of space junk flying about
and the little NVIDIA guy in the shiny black jacket updates his chips every year.

I'll stick to day trading Spacex.
Thats saved me 2 pints, thanks Justin
 
Investing can be for each of us to pay for care homes, sure, but
arguing the toss and criticising each other about how the GOVERNMENT should pay for it, is not.
SImple enough?

Don't try investing, you lot, you're too thick!


#488 is where the trollling started.


you mean the post where @JohnD say "You didn't mention how it would be funded." to @motorbiking

That's your definition of trolling?

Okay.

Conveniently ignore the post prior at #480 where you introduce the subject. I need no more gaslight than already provided by the s.fix trolls, thanks. I agree the topic of investments can include care homes or any other area of private investment: a case can be made that the private sector has inflated costs and reduced the quality of care since the sector was introduced by Maggie T. and AB is correct to consider bringing the care home system back under public ownership in the form of council care. This will be seen as a move towards a Socialist ideal by opponents of this way of doing things but a method of funding care homes must be found and maintained without reducing famillies to putting their life savings into a year or two of care at the end of a family members life.

Hardly anyone was reading your other thread as the s.fix trolls hunt in a pack so it takes one to start a thread for the others to become invested in the topic. Carry on...
 
Should you invest in Spacex?
2 Trillion dollar thing. Data centres in space
Power is free from the sun, and it's cold, so cooling is easy.
Elon says.

I think he's fibbing.
Date centres use gigawatts of power. A big one would be a Terawatt.
Solar incident power is of the order of 1kW per square metre. It's a little easier in space because of the exposure and no night. The lack of air makes almost no difference.
So you need a solar collector . Lets assume they're 100% efficient - they're only about 20% really.
1TW is 10^12 W
10^12 / 10^3 square metres
That's 10^9 sq m.
You could use 10 panels 10^4m x 10^4m
That's 10 square panels, ten kilometers on each side.
Origami anyone?

Then there's the cooling.
Yes it's cold in space. 3Kelvin or so.
But the "it" is almost nothing. It's a vacuum!
So there's no stuff to heat up and move away from your data centre carrying the heat.
You can use radiative emission. That sounds good, because , if you remember skule fiziks, it's proportional to the temperature difference to the power 4. Do the names Stefan & Boltzmann ring a bell?
Your chips run at say 50C which is 323K
Then you plug in the constant, which is 5.670367 × 10⁻⁸ W·m⁻²·K⁻⁴
Spot the pesky -8 in there.
It comes out that you need about 1 order of magnitude less for the size of radiator to cool the thing, than you needed to get the energy.
And they're really heavy, compared with solar collectors. You need a couple of thousand square kilometers of the things, all got there by rocket.

Oh, and the data centre itself would be pretty big and heavy. Maybe 100 square kilometers, but they're definitely three dimensional.

Then there's the maintenance problem
and bits of space junk flying about
and the little NVIDIA guy in the shiny black jacket updates his chips every year.

I'll stick to day trading Spacex.
There may soon be enough tech in space to make a business case for a "space cloud". The main issue you have identified is maintaining the facility. MTBF is 40k-100k hours per server. Assuming a 1GW DC, thats one server per 2 hours on average. Some failures will have greater impact. So yeah, the idea needs work.
 
Last edited:
By page 2 of this thread advice was given to invest in marrow bone, dingys and a disused railway station. It’s diynot, not the FT
 
What's WRP?
I think its this but its scary stuff to me.

Screenshot_20261103_171404_Chrome.jpg
 
Back
Top