3 ms·

Why, because on board storage is too small and the compute nodes are underpowered? And that can't ever change? A reasonable person doesn't understand technology

by asdff 4mo ago

Why, because on board storage is too small and the compute nodes are underpowered? And that can't ever change? A reasonable person doesn't understand technology usually. That is increasingly an understanding left to the wizard class. I mean people make clusters out of raspberry pis and minipcs.

dgellow 4mo ago
There is nothing magical here, you definitely don’t need to be a wizard to understand the hardware necessary for AI inference. You can make nice little clusters with rpis, yes, there is nothing magical about that, it’s pretty much baremetal 101. But no, you cannot run any meaningful inference on that cluster. Maybe look at what is inside a datacenters, the amount of power required is very large, and the hardware to run the inference + network isn’t small. Then try to see how much sending that to space cost
wasabi991011 4mo ago
> Why, because on board storage is too small and the compute nodes are underpowered? Yes > And that can't ever change? It can, but but not for free. As the comment earlier in the thread was referring to, more computing power needs more cooling, and cooling in space is hard. > A reasonable person doesn't understand technology usually. What? Do you mean a layperson? Why does that matter when discussing the feasibility of space-based AI datacenters? > That is increasingly an understanding left to the wizard class. No, you can get there with a bachelor's degree in a relevant subject. Or just reading informative news sources. > I mean people make clusters out of raspberry pis and minipcs. So? What does that have to do with anything?
frollogaston 4mo ago
This only makes sense if the compute nodes only become more powerful in space in the future and not also on the ground.
- asdff 4mo ago
  The big advantage is if you put it in space, only people with space based capabilities can mess with it. If you build it on the ground, anyone can drive a truck full of explosives into it. Might not matter for most applications, but there are some where it absolutely matters I am sure.
  - frollogaston 4mo ago
    They can cross that bridge when they get there. Currently they're at the stage where the energy requirement is simply too much to put in space, it's not even close, and they will work on reducing that whether or not it goes into space.

by asdff 4mo ago

dgellow 4mo ago

There is nothing magical here, you definitely don’t need to be a wizard to understand the hardware necessary for AI inference. You can make nice little clusters with rpis, yes, there is nothing magical about that, it’s pretty much baremetal 101. But no, you cannot run any meaningful inference on that cluster. Maybe look at what is inside a datacenters, the amount of power required is very large, and the hardware to run the inference + network isn’t small. Then try to see how much sending that to space cost

wasabi991011 4mo ago

> Why, because on board storage is too small and the compute nodes are underpowered? Yes > And that can't ever change? It can, but but not for free. As the comment earlier in the thread was referring to, more computing power needs more cooling, and cooling in space is hard. > A reasonable person doesn't understand technology usually. What? Do you mean a layperson? Why does that matter when discussing the feasibility of space-based AI datacenters? > That is increasingly an understanding left to the wizard class. No, you can get there with a bachelor's degree in a relevant subject. Or just reading informative news sources. > I mean people make clusters out of raspberry pis and minipcs. So? What does that have to do with anything?

frollogaston 4mo ago

This only makes sense if the compute nodes only become more powerful in space in the future and not also on the ground.

asdff 4mo ago
The big advantage is if you put it in space, only people with space based capabilities can mess with it. If you build it on the ground, anyone can drive a truck full of explosives into it. Might not matter for most applications, but there are some where it absolutely matters I am sure.
- frollogaston 4mo ago
  They can cross that bridge when they get there. Currently they're at the stage where the energy requirement is simply too much to put in space, it's not even close, and they will work on reducing that whether or not it goes into space.