Power Into Energy -> CPU edition.

[lavahot]

Patrick is partially right when he says that not all of the energy is transformed into heat. While resistors, capacitors, and transistors are the main heat sources in a computer, not all of the energy supplied to the machine is transformed into heat through the calculative processes. There are parts that transform the electrical energy into mechanical energy, like hard disks, optical drives, and fans. And then there are LEDs, which don't take up much energy, but transform it into light energy.

So my point is this: if you put a computer in a sealed environment, then ran it for an amount of time while measuring it's electricity consumption, the temperature change in the sealed environment would not account for the total energy consumed by the computer due to the other types of energy released.

[computoman]

I do know that in my mancave(home office), if I have a couple of desktops and servers running, you do not have to run the central heating system in the house if no one else is home except during extreme cold weather.

[revision3fan]

In 2000, there was a TV series called Level 9. The main computer room was called "the Fridge".

[fishtoprecords]

put a computer in a sealed environment, then ran it for an amount of time while measuring it's electricity consumption, the temperature change in the sealed environment would not account for the total energy consumed by the computer due to the other types of energy released.

What other types of energy are released?

This is absurd. Measure the flow of energy, convert to watts, add in proper power factor, multiple by 3.41 and you have BTUs.

There is nothing else going on.

How do you propose to make this "sealed environment" that has no energy leakage?

[artsmart]

OMG, I'm amazed that this whole issue is controversial. Every last joule of energy that goes through the power cord of a PC ultimately ends up as heat. Period.

Yes, energy can be stored into capacitors, but after the PC is shut down, that energy eventually discharges, and the result is heat. Yes, energy can be stored into the magnetic field surrounding an inductor, but cut off the power, and that field quickly collapses by forming eddy currents in nearby conductors, and that energy ends up as heat. Laptop batteries can be charged, thereby temporarily storing energy in chemical form. But once that energy from the battery is used, it ends up as heat.

Someone mentioned LEDs emitting light. Yes, they emit light, but once that light hits a wall, a shirt, or your retina, even that photonic energy ends up as heat. Even if the LED were used to provide artificial sunlight to grow a plant, once that plant dies and decomposes, it gives up the stored energy in the form of heat. There is absolutely no way to escape the ultimate form of the energy being heat. Every last joule of it.

On the separate issue of power vs. energy, yes, power is energy per unit of time. Or you could say energy is power consumed during a specific period of time. A kilowatt of power being consumed for an hour is a kilowatt hour of energy consumed. And that kWh of energy is equivalent to 3,413 BTUs.

I'm not an electrical engineer, but rather a chemical engineer (BS 1976, MS 1977, both from Georgia Tech). One of the most critical aspects of chemical engineering is to understand how an energy balance works. Been doing them for 38 years now.

[fishtoprecords]

OMG, I'm amazed that this whole issue is controversial. Every last joule of energy that goes through the power cord of a PC ultimately ends up as heat. Period.

Right.

Joule is just the standard system (metric or SI) energy unit.

9.4782×10−4 BTU (British thermal unit)

In the US, we are stuck using BTU rather than joule. So our furnaces and air conditioning systems are rated in BTU (or for A/C, tons).

People in this thread and others seem to be confusing "efficiency" with usage. A 1000 watt power supply may be rated at 80% efficiency, which just means that for every 1000 watts in, the power supply itself only wastes 200 watts of power (or 682 BTU) leaving up to 800 watts for the motherboard, CPU, disks, etc. to turn into heat.

[dimf]

Apart from the fact that ArtSmart is 100% right I am more interested in Big Mikes reply (in episode 87).
How come someone manages to explain that there is a difference between power and energy and then fails to explain what happens to the power _not_ dissipated as heat by the PSU? Could it become kinetic energy pushing the computer across the floor? or maybe potential energy making the computer levitate?

This should be part of everybodys physics education, at least it is in the metric parts of the world.

Also, isn't the wattage specification of a PSU how much it can deliver - not how much it consumes from the mains. A 1000W PSU fully loaded with a 80% efficiency uses 1250W. Simple mathematics from 5th grade.

We should make a really funny TV show where people are asked questions from 5th grade and miserably failing to get the answers. Oh wait, it has already been done and it is not very funny but scary how little american people actually know.

[revision3fan]

The power not dissipated by the power supply as heat is transferred to the various components; drive motors, processors, memory, etc.

[bigshotprof]

In 2000, there was a TV series called Level 9. The main computer room was called "the Fridge".

Level 9. Good Times. Could a guy really have the name Max Martini?