Thread 16781294

Computation in theory should consume zero (or almost zero) power.
Yet it doesn't in practice.
Why is that?

>>16781294 (OP)
Why do you think it should consume almost zero power?

>>16781336
Actually it can consume exactly zero power - in reversible computing.
But even without reversible computing actual theoretical energy needs are orders of magnitude lower of what we have currently: https://en.wikipedia.org/wiki/Landauer%27s_principle

>>16781340
>dissipates a minimum amount of heat to its surroundings.
Minimum is not zero. Your own link proves you wrong in the first sentences. There is no perfect energy use in existance. I can't think of a single process in nature that has a 100% efficiency.

You can lower energy consumption by turning your pc off now.

>>16781354
Celestial object in orbit loses almost zero energy during its movement through vacuum of space.

>>16781361
Almost zero is not zero....... and your example depends heavily on the speed of these objects.

>>16781361
Because if gravity isn't acting in it there's no difference between an object moving fast and objects moving towards it fast. It may as well be still

>>16781421
Depends on how fast we are talking about. Near c the cosmic backround radiation becomes so hard that it's like running headfirst into a brickwall.

>>16781361
Tital lock

>>16781354
>I can't think of a single process in nature that has a 100% efficiency.
Isn't matter-antimatter annihilation a "100%" efficient source of energy?

>>16781361
yeah and you're not extracting any work out of it

>>16781519
That's the point.
Computation isn't work physically speaking.

>>16781294 (OP)
Because the circuitry gets heated? Duh

>>16781294 (OP)
Silicon technology is shit.

>>16781537
It is if you want a result out of it. Nature is losslessly computing gibberish at gorillion operations a second constantly.

>>16781486
No

>>16781537
On quantum level you start loosing energy while using electricity to power your computer... you can't built a computer with 100% efficience because quantum mechanics didn't allow it. It's not possible in our universe. You will always have some energy loss in the process.

>>16781604
Why would you simply go on the internet and post wrong answers to questions in subjects you know nothing about?

>>16781616
It's just not possible to build a computer with 100% efficience. You would need to be able to use every kind of energy that is created/changed in the process.
How do you want to make 100% of heat usable? The magnetic field your pc creates, how do you want to use 100% of it's energy? It's not possible. You will always have some energy loss somewhere.

>>16781354
>I can't think of a single process in nature that has a 100% efficiency.
if a photon moves through empty space for a planck meter in a plack second does it shift frequency by a planck hertz (lose energy)? photon moving a planck meter is computation

>>16781679
What computation has been performed by a lone photon moving in a straight line?

God didn't make the computer. You did.

>>16781679
That's stupid. To get information out of it, you need to measure it = losing energy

>>16781361
Yea, I am sure that is why all the planets are constantly being blasted with massive amounts of solar radiation because celestial objects don't lose energy.

>>16781294 (OP)
>Instant entropy
ok

>>16781336
>Why do you think it should consume almost zero power?
https://en.wikipedia.org/wiki/Negentropy

In 1953, Léon Brillouin derived a general equation[17] stating that the changing of an information bit value requires at least
k T ln2

>>16781870
the computation of the end state of a photon moving in a straight line?
for example, if you want to compute the state of coffee in a mug you can put coffee in a mug, you are then looking at the final result of the computation. I don't understand why this is tripping people up.
you can try and make a digital computation of what a storm develops as, but you have way higher fidelity from actually running the storm on bare metal. you have worse access to all data points but you have max fidelity.
or maybe simulate how much a racing tire wears from running on a race circuit with a racing car that has x amount of horse power. you can simulate it, or install the tire on the car, run the circuit, and analyze the end computation (result, tire itself).

>>16782626
>if you want to compute the state of coffee in a mug you can put coffee in a mug, you are then looking at the final result of the computation

You still have to look to see the result. Look = using energy
Computing without measuring is retarded concept. How do you know that something is computing without measuring it. And measuring always means using energy.

Even our brains need 20 watts, with the rest of the body serving as the modular PSU.

>>16781679
Why the Planck units in particular? Why would it matter if the photon was moving for smaller or larger distances relative to the Planck unit? Planck lengths/areas/times aren't a lower bound on the possible states of material reality. They are a lower bound on the size in which classical physics is useful.

>>16782626
Observation of the state of the photon requires energy.

>>16781294 (OP)
>omputation in theory should consume zero (or almost zero) power
According to what theory, jackass? All thermodynamic processes require a source of energy, and the mass-energy equivalency theorem directly implies that thermodynamic processes necessarily involve the conversion of mass to energy or vice versa.

>>16781361
Infinitesimal =/= zero. You just got filtered by 17th century math.

>>16783787
>They are a lower bound on the size in which classical physics is useful.
So the mass of a grain of dust is about the limit at which the concept of mass makes sense? Thats wild

>>16781427
"Run Achilles, run!", the unreachable Tortoise taunted.

OP. You have started the thread with a question. “Why am I stupid?”

>>16781294 (OP)
None of our computers compute in theory, they all do crazy shit with charging and discharging microscopic capacitors and all conductors generate magnetic fields when carrying a current and little metalization filaments in chips are shit conductors not superconductors. It's all barely good enough to work but not 100 percent efficient.

>>16784517
> So the mass of a grain of dust is about the limit at which the concept of mass makes sense? Thats wild.

I'm an engineer, not a physicist. In my experience, the concept of mass stops making sense as soon as you need precision. It's a fantastic way to get you into the ballpark, but as soon as you're talking about fitting a 1m wide object into a 1 cm wide box based on measurements you observe from 100 km away, you've confused the forest for the trees.