Right, but like… whatever you’re doing in space is going to be more cost effective to do on earth. Not to mention the insane amount of energy lost to the atmosphere
Unless you really need to optimise for land use. An arbitrarily large solar array in space could transmit to a fairly small collector in the surface.
As for losing power to atmospheric attenuation, high frequency microwaves will pass right through most everything that would scatter visible light. Clouds, dust, etc wouldn’t really impede it.
I won’t say it’s not a silly idea, because it is. It’s fun to think about though.
musk wants datacenters in space. which makes sense, 24/7 sunlight and no transmission of power is grand; but I do wonder about the shielding and moving the data back and forth.
yeah had a whole convo with a neighbor about how much cooling tech the ISS depends on.
at least it won’t need separate water/ammonia loop setup like the ISS. radiators are pretty figured out. I just can’t see how they make it economical with all the launch and space logistics - and don’t get me wrong spacex can deliver to orbit - but can they make it profitable?
also, where’s the grunt for this supposed AI cloud gonna come from? what chips can survive for 1000s of hours of compute in that environment? and from what he’s said (24/7 sunlight) are they thinking lagrange points or what? also xmit/receive of massive amounts of data would need to be crucial to making it work and we got none of that infrastructure…
Right, but like… whatever you’re doing in space is going to be more cost effective to do on earth. Not to mention the insane amount of energy lost to the atmosphere
Energy loss for wireless energy transmission is actually surprisingly low. Here is an example of 80% efficiency over 1 kilometer: https://ieeexplore.ieee.org/document/1123672
…
Unless you really need to optimise for land use. An arbitrarily large solar array in space could transmit to a fairly small collector in the surface.
As for losing power to atmospheric attenuation, high frequency microwaves will pass right through most everything that would scatter visible light. Clouds, dust, etc wouldn’t really impede it.
I won’t say it’s not a silly idea, because it is. It’s fun to think about though.
You could also have a constellation of satellites with area greater than the surface of the earth. It’s not that silly of an idea.
Big solar panel
musk wants datacenters in space. which makes sense, 24/7 sunlight and no transmission of power is grand; but I do wonder about the shielding and moving the data back and forth.
It’s nonsense, for cooling reasons alone
yeah had a whole convo with a neighbor about how much cooling tech the ISS depends on.
at least it won’t need separate water/ammonia loop setup like the ISS. radiators are pretty figured out. I just can’t see how they make it economical with all the launch and space logistics - and don’t get me wrong spacex can deliver to orbit - but can they make it profitable?
also, where’s the grunt for this supposed AI cloud gonna come from? what chips can survive for 1000s of hours of compute in that environment? and from what he’s said (24/7 sunlight) are they thinking lagrange points or what? also xmit/receive of massive amounts of data would need to be crucial to making it work and we got none of that infrastructure…
all leads me to think his ketamine is showing.
yeah but imagine you can put the plant and all the pollution on an asteroid or something.