[quesera]

Using a cryptominer or GP computer at full tilt for the purpose of making heat is less than 100% efficient, because some of the input power gets used on computing.

A pure resistive heater is also less than 100% efficient, but it's very very close.

[jffry]

Where does the energy "used on computing" go? The power consumption of a CPU is dissipated as heat.

[quesera]

So, I think I understand this view from the thermodynamics sense. First law and all.

But by this construction, are all systems which take energy as input (and do not convert it to another form for storage) de facto 100% efficient?

When we talk about a (very old) furnace being 75% efficient at turning the chemical energy of a fossil fuel into heat energy, is the 25% loss purely combustion byproducts with some inherent chemical energy plus some non-combusted fuel?

[jffry]

In this case for heating, efficiency is how much of the energy input gets converted into heat in the space you're heating.

So for a resistive electric heater, you're dissipating all of the energy as heat (minor quibbles about electromagnetic radiation or status LEDs aside). The same is true for your computer.

Burning hydrocarbons have two main sources of energy loss: incomplete combustion, and energy carried away by exhaust gases. Obviously furnace design influences both of these - the US federal minimum is 78%, but high-end furnaces with all the tricks can get over 90% per https://www.energy.gov/energysaver/furnaces-and-boilers

Heat pumps are the third major category, and they are more than 100% efficient because they are using their energy to steal a larger amount of thermal energy from the outdoors air and move it to inside your home (and much of the electric energy they consume is eventually discharged as heat inside your home as well).

[amluto]

Your computer is not doing work (in the thermodynamic sense) that gets stored anywhere in the computer, and it is not transmitting a significant amount of energy out the Ethernet port either. (It is moving some energy out the port, copper or fiber, but it’s likely receiving an almost equal amount back, and the power is question is negligible.)

[quesera]

I think what you're saying is that Max Planck knew a thing or two.

[amluto]

Huh? Maybe Rolf Landauer.

But this is really just conservation of energy or the first law of thermodynamics, and those predate Planck.