Likely the intended meaning here is that the practicality of space data centers goes against the physical realities of operating in space. The single most prevalent issue with operating anything in space is heat dissipation in that the only method of doing so is via radiation of heat, which is very slow. Meanwhile, the latest Nvidia reference architectures convert such ungodly amounts of power into heat (and occasionally higher share prices) that they call for water cooling and extensive heat-exchange plant.

Even if one got the the economics of launching/connecting GPU racks into space into negligable territory and made great use of the abundent solar energy, the heat generated (and in space retained) by this equipment would prevent running it at 100% utilization as it does in terrestrial facilities.

In addition to each rack worth of equipment you'd need to achieve enough heat sink surface area to match the heat dissipation capabilities of water-cooled systems via radiation alone.

Not physics defying, just economically questionable.

The main benefits to being in space are making solar more reliable and no need to buy real estate or get permits.

Everything else is harder. Cooling is possible but heavy compared to solar, the lifetimes of the computer hardware will probably be lower in space, and will be unserviceable. The launch cost would have to be very low, and the mean time between failure high before I think it would make any economical sense.

It would take a heck of a lot of launches to get a terrestrial datacenter worth of compute, cooling and solar in orbit, and even if you ship redundant parts, it would be hard to get equivalent lifetimes without the ability to have service technicians doing maintenance.

Their viability is what I called physics-defying. Without some drastic changes to our current level of technology, the added costs of putting something in space along with the complexities of powering, cooling, and maintaining it once it's there is just too much to overcome the alternative of just building it on Earth.

Cooling.

Radiative cooling is the only option, and it basically sucks vs any option you could use on earth.

Second, ai chips have a fixed economic life beyond which you want to replace them with better chips because the cost of running them starts to outpaxe the profit they can generate. This is probably like 2-3 years but the math of doing this in space may be very different. But you can't upgrade space based data centers nearly as easily as a terrestrial data center.

More details from a guy who has thought this through https://taranis.ie/datacenters-in-space-are-a-terrible-horri...

How do you cool them? Getting rid of heat is one of the number one challenges on the ISS.

Without evaporation and convection, getting rid of heat is a bitch in space.