First, does a computer chip require an atmosphere to run? In other words, it can run in a vacuum, right?

 

Second, I know there is an upper limit to acceptable temperatures for chips to run correctly in (hence heatsincs and fans in computers), but is there a lower limit?

Yes it can run in a vacum.

 

I doubt it has a lower limit.

 

I've seen computers cooled by nitrogen and almost frozen, and they still work.

Suprisingly my computer manual recommends no operating temperatures below 10C but I too have seen the liquid nitrogen cooling systems working in the video of the world record 5GHz project... but then that processor was kicking out a lot of heat, but it still woulda lost it effectively immediately.

 

Running a computer in a vacuum would work, but there'd be little heat loss so it could over heat.

There are physics formulas for that of course quite advanced.

I too have seen the liquid nitrogen cooling systems

 

Well then, that answers my question - I was considering experimenting with one, and was wondering if this would kill things. Apparently not. Thanks for the help.

While I'm on the subject, do cooling systems have to be on the chip? In particular, could the chip be in a supercold atmosphere (like with a liquid nitro cooling system) without direct contact with the cooling system, without running the risk of overheating?

they can work in a vacuum if they`ve been tested for that, the resin outer coating can contain air pockets on cheaper chips (that`s why some explode when shorted out or over heated) the expand and `POP`, a Vacuum can and will have the same effect.

 

as for operating temp, there IS a lower limit (depending on the substrate/ chip type) mostly it`s important with analog ICs as they deal with Voltages rather than binary ON/OFFs at frequency.

semicons are suceptible to temp where voltage is concerned and so an offset with temp reference is usualy required to compensate.