Timeline for How many bits are needed to simulate the universe?
Current License: CC BY-SA 3.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 2, 2015 at 10:18 | comment | added | kutschkem | Are you sure the wave-function has this much entropy? | |
| Dec 29, 2012 at 14:38 | comment | added | Anixx | This is not a precise answer. It is true only if you are speaking about a storage that can operate with variables of no less than 1 bit size. In that case you need $10^{80}$ such variables. But the majority of that space is wasted, because in reality each variable does not need a FULL bit. That is by compressing several such variables into one bit you'll need much less storage. Unfortunately the classical devices cannot independently manipulate quantities of information less than 1 bit. | |
| Aug 3, 2012 at 15:26 | vote | accept | z5h | ||
| Aug 1, 2012 at 7:02 | comment | added | Ron Maimon | @Raskolnikov: Quantum monte-carlo only works well for bosonic systems, it is very hard to do with fermions, and this is the "sign problem". It is extraordinarily depressing that something as simple and dumb as a large atom is so hard to simulate. | |
| Apr 21, 2011 at 22:05 | comment | added | z5h | I think the key is that in order for us to care, the events must be measurable. If they are not all measurable, we don't need all of them in our simulation. see here: en.wikipedia.org/wiki/Infrared_divergence | |
| Apr 21, 2011 at 21:57 | comment | added | Lagerbaer | Couldn't you have an arbitrary number of virtual particles albeit on a small timescale? | |
| Apr 21, 2011 at 21:53 | comment | added | z5h | "in a black body, for example, you can have an infinite number of photons". If we assume finite energy and finite space in the universe, and we can encode a 1 bit of data per photon, this disagrees with the Bekenstein bound. The other option is that although we have infinite photons, we cannot decode the information stored. So we can discard them from our simulation. no? | |
| Apr 20, 2011 at 8:29 | comment | added | Raskolnikov | Hah, I answered my own question. I iz a winnah! | |
| Apr 20, 2011 at 8:27 | comment | added | Raskolnikov | This makes me wonder, isn't there a stochastic (Monte Carlo) way of simulating quantum mechanical systems akin to what is done when solving SDE and Markov processes and such? | |
| Apr 20, 2011 at 3:33 | comment | added | Lagerbaer | This is the nasty thing about quantum mechanics. You have to store the value of the wavefunction for each of the possible combinations of your $\vec{x}_i$. You have $n$ for the first, $n$ for the second, $n$ for the third... so in total $n^m$. | |
| Apr 20, 2011 at 3:17 | comment | added | yayu | @lagerbaer basic question: if you wish to encode an m parameter function over n points, why would it be $n^m$? | |
| Apr 20, 2011 at 0:31 | history | answered | Lagerbaer | CC BY-SA 3.0 |