Right, that's the point! The low pressure would make it very very inefficient, and you would actually need a heat exchange system far larger than earth reactors

Right, that's the point! The low pressure would make it very very inefficient, and you would actually need a heat exchange system far larger than earth reactors. We use liquid cooling in our PCs because a smaller heat sink can pump out more heat if it's going into liquid (for a fixed flow rate). This is why liquid cooling for CPU/GPU setups is quiet and effective: the liquid flows quietly and carries heat to a BIG, slow fan. Your fan and heat sink would be even bigger/faster at Mars atmospheric densities just to cool your puny CPU. Doing that at scale with no water cooling for your reactor would be absurdly materials intensive. That's all for a typical high-power reactor setup. You can, of course, have a low-power setup with nuclear power. RTGs [1] use this approach, though they mainly rely on decay heat rather than self-sustaining reactions, which is just as well because they are also thermodynamically limited by the rate at which they can radiate heat away. One addendum I should have noted: the efficiency of a thermal engine on mars would actually be perfectly reasonable for a small engine (it's nice and cold there, perfect for high efficiencies). The issue is that you can't dump the heat fast enough, so your heat sink will get hotter, trashing your efficiency and reducing your cooling capacity (both of which exert downward pressure on your final power output in a serious way). [1] https://en.wikipedia.org/wiki/Radioisotope_thermoelectric_generator https://en.wikipedia.org/wiki/Radioisotope_thermoelectric_ge...