3 ms·
> that simulating quantum mechanics is so computationally expensive that it isn’t done unless we’re looking at it This is a common misconception. It's more dif
by gus_massa 8y ago
> that simulating quantum mechanics is so computationally expensive that it isn’t done unless we’re looking at it
This is a common misconception. It's more difficult to simulate a quantum system than a classical system. To simulate a classic system you have to simulate one posible state. To simulate a quantum system you must simulate all possible path.
This is the advantage of quantum computers. With the hardware for only N qbits, you can calculate 2^N states (for some very specific problems that are quantum-computer-friendly).
- amelius 8y agoI'm by no means an expert, but how about using Pilot Wave Theory to do the simulation? As I understand, you wouldn't need to simulate an exponential number of states, you just simulate the pilot wave. See also (very accessible): https://www.quora.com/Why-dont-more-physicists-subscribe-to-pilot-wave-theory https://www.quora.com/Why-dont-more-physicists-subscribe-to-...
- gus_massa 8y agoYou must simulate the value of the wave at every point of the universe, or at least a grid that has enough points to have the correct precision. It's even more work. Imagine simulating all the waves in all the ocean in the Earth, but worse because there are some nonlocality effects. You can probably do some kind of smart representation of the wave, were you decompose it in some smart functional base of smooth waves. With a very optimized representation the calculations would be equivalent to the usual quantum mechanics calculations. With a not smart enough representation you would get something that needs more calculations.
- amelius 8y ago> You must simulate the value of the wave at every point of the universe Maybe I'm not getting it, but that still sounds simpler than simulating an exponential number of states for the entire universe.
- gus_massa 8y agoTake a look at https://en.wikipedia.org/wiki/Pilot_wave_theory#Mathematical_formulation_for_multiple_particles https://en.wikipedia.org/wiki/Pilot_wave_theory#Mathematical... The problem is that if you have five particles then the function ψ(r1, r2, r3, r4, r5, t) is a function on 3*5+1=15 coordinates. If we assume that for each particle we use a 100x100x100=1000000 grid, then for the 5 particles we need a grid of (1000000)^5=1000000000000000000000000000000 points. The space were the wave is defined grows exponentially. (As a rule of thumb, with a current notebook, you can do up to 1000000000 or 1000000000000 calculations in a few seconds/minutes/hours.)
- dr_dshiv 8y agoI was wondering how this article would read from a Bohmian perspective. It would seem to be an alternative to creating the universe as a result of perceptions in the modern era. (Damn Copenhagen interpretation) I was also wondering about entropy as the arrow of time -- positrons are just as subject to entropy, no?
- kgwgk 8y agoThe equations are not simpler, there is no “just”.