Gorgeous Simulation Creates First Realistic Virtual Universe

[Rhykker]

Gorgeous Simulation Creates First Realistic Virtual Universe


Astronomers have developed the first realistic virtual universe in a computer simulation that recreates 13 billion years of cosmic evolution.

In the past, attempts to simulate the evolution of our universe were met with technological hurdles that resulted in either low-resolution or small-scale simulations. But a new simulation named "Illustris" has done the unprecedented and given us the first detailed view of a realistic, virtual universe that closely parallels our own.

Illustris simulates an evolving universe within a cube that measures 350 million light-years on a side, beginning 12 million years after the Big Bang and stretching on for 13 billion years. Astronomers counted more than 41,000 galaxies in cube at present day, with a realistic mix of spiral and elliptical galaxies, galaxy clusters, and accurate depictions of the chemistries of individual galaxies.

It took five years to develop the Illustris program, and the simulation took three months to render, using 8,000 CPUs running in parallel. Had the team used an average desktop PC, the calculations would have taken more than 2,000 years to complete.



Illustris was a joint effort between researchers at several institutions, including the Harvard-Smithsonian Center for Astrophysics and the Heidelberg Institute for Theoretical Studies in Germany. The results are being reported in the May 8th issue of the journal Nature.

High-definition video and images of the simulation can be found on the official website [http://www.illustris-project.org/]. Simulated or not, the cosmos is filled with beauty.


Source: Harvard-Smithsonian Center for Astrophysics [http://www.cfa.harvard.edu/news/2014-10]

Permalink

 

[kanetsb]

Now here's a brain-fck for ya:
Imagine this, being simulated on a quantum super-computer (or maybe even something better?), many years in the future... but the calculation scale is sub-atomic instead of semi-galactic.

Now... Since this would be a calculated simulation, the internal time flow would be discrete ('frame-by-frame') and as such, certain effects would occur (if you ever fell through the floor in some game - that's the reason, time is not continuous in the physics simulation). One of such effects would be stuff sometimes going through stuff, or as we call it - quantum tunneling, something we are observing in our sim...I mean universe...

 

[Hagi]

Now here's a brain-fck for ya:
Imagine this, being simulated on a quantum super-computer (or maybe even something better?), many years in the future... but the calculation scale is sub-atomic instead of semi-galactic.

Now... Since this would be a calculated simulation, the internal time flow would be discrete ('frame-by-frame') and as such, certain effects would occur (if you ever fell through the floor in some game - that's the reason, time is not continuous in the physics simulation). One of such effects would be stuff sometimes going through stuff, or as we call it - quantum tunneling, something we are observing in our sim...I mean universe...

Only problem is that in order to simulate something on a sub-atomic level you'd need something at least as big as the thing you're simulating. As in, whatever physical medium is used to represent said virtual sub-atomic particle has to be a sub-atomic particle or bigger in order to be able to contain all the necessary information. In layman's terms, you can't simulate 8 gigs of RAM using 4 gigs of RAM, that's impossible.

As such said simulation would have to be run on a computer that was at least as big as the universe it's simulating. And since we'll probably never be able to perfect technology at 100% efficiency, where every single particle involved is being used as much as it can be, probably even bigger than that.

Considering we don't even know just how big our universe is, we can't even observe all of it. That theoretical computer simulating our universe would have to be so insanely large that it really doesn't make any sense whatsoever anymore. Certainly, I can't truly disprove it but a supercomputer that's at least as big as our universe? An extraordinary claim, requiring extraordinary proof vastly more than our the simple fact that we don't fully understand quantum mechanics.

And it's still very much a question whether or not a quantum computer would actually do this sort of thing faster. They're not magical things that make everything go insanely fast. It's simply a different way to calculate, much more efficient at certain calculations and much less so at others.

 

[drakonz]

Now here's a brain-fck for ya:
Imagine this, being simulated on a quantum super-computer (or maybe even something better?), many years in the future... but the calculation scale is sub-atomic instead of semi-galactic.

Now... Since this would be a calculated simulation, the internal time flow would be discrete ('frame-by-frame') and as such, certain effects would occur (if you ever fell through the floor in some game - that's the reason, time is not continuous in the physics simulation). One of such effects would be stuff sometimes going through stuff, or as we call it - quantum tunneling, something we are observing in our sim...I mean universe...

Considering we don't even know just how big our universe is, we can't even observe all of it. That theoretical computer simulating our universe would have to be so insanely large that it really doesn't make any sense whatsoever anymore. Certainly, I can't truly disprove it but a supercomputer that's at least as big as our universe? An extraordinary claim, requiring extraordinary proof vastly more than our the simple fact that we don't fully understand quantum mechanics..

and how are you sure that our laws of physics are same as the universe that is simulating ours? things like this cant rly be disproved in similar way as idea of god you can always come up with reason why it would be posible.

 

[Hagi]

and how are you sure that our laws of physics are same as the universe that is simulating ours? theories like this cant rly be disproved in similar way as idea of god you can always come up with reason why it would be posible.

As I said in my post, I can't disprove it.

I can however claim it to be extraordinarily unlikely with an explanation as to why that is. And every additional reason that needs to be invented for why a thing is possible means it requires additional assumptions in order to be believed and provides additional doubts, weakening the entire theory.

 

[kanetsb]

BTW, this is not a new problem... It's an old philosophical and scientific conundrum described here for instance: http://en.wikipedia.org/wiki/Simulated_universe - many have tried figuring this out before. And obviously... The Matrix needs to be mentioned and quickly set aside.

It's just something that "made me think" about this again.

Only problem is that in order to simulate something on a sub-atomic level you'd need something at least as big as the thing you're simulating.

With regards to that problem - let's say we're trying to simulate on a quantum level - just imagine that the computer used, is based on some kind of futuristic super-string computation system... oh wait... BOOO!!! - https://www.youtube.com/watch?v=bp4NkItgf0E

 

[Vigormortis]

and how are you sure that our laws of physics are same as the universe that is simulating ours? things like this cant rly be disproved in similar way as idea of god you can always come up with reason why it would be posible.

All of these "reasons" amount to nothing if you can't prove that either scenario is true.

The onus of proof is on the person making the positive claim. And until that person can provide such proof; specifically quantifiable, demonstrable, repeatable evidence; than the default position is to not accept the claim as true. It's not up to others to disprove these claims. You can't prove a negative.

"Extraordinary claims require extraordinary evidence." A truism in the purest sense.

 

[Hagi]

Only problem is that in order to simulate something on a sub-atomic level you'd need something at least as big as the thing you're simulating.

With regards to that problem - let's say we're trying to simulate on a quantum level - just imagine that the computer used, is based on some kind of futuristic super-string computation system... oh wait... BOOO!!! - https://www.youtube.com/watch?v=bp4NkItgf0E

It doesn't matter how fast the computer is. It's not about speed. It doesn't matter what computational system it uses. It's not about calculations.

It's about information.

The data of sub-atomic particles would have to be stored, all of it's information. Everything. The only thing that can contain all the information that makes up a sub-atomic particle is a sub-atomic particle or something bigger.

The memory of your super-computer, regardless of what processor it used, regardless of how that processor did it's calculations, regardless of how fast that processor was, would have to be at least as large the universe it was simulating on a sub-atomic level in order to contain all the information contained within that simulated universe on a sub-atomic level.

That's just the memory, it has to be that big. Can't store 2 bits of information within a single bit. Can't store 2 sub-atomic particles of information within a single one. Not even talking about the rest of the architecture required for such a computer. Just the memory.

If our universe is a simulation then it's stored on something that's larger than our universe is.

 

[drakonz]

Only problem is that in order to simulate something on a sub-atomic level you'd need something at least as big as the thing you're simulating.

With regards to that problem - let's say we're trying to simulate on a quantum level - just imagine that the computer used, is based on some kind of futuristic super-string computation system... oh wait... BOOO!!! - https://www.youtube.com/watch?v=bp4NkItgf0E

It doesn't matter how fast the computer is. It's not about speed. It doesn't matter what computational system it uses. It's not about calculations.

It's about information.

The data of sub-atomic particles would have to be stored, all of it's information. Everything. The only thing that can contain all the information that makes up a sub-atomic particle is a sub-atomic particle or something bigger.

The memory of your super-computer, regardless of what processor it used, regardless of how that processor did it's calculations, regardless of how fast that processor was, would have to be at least as large the universe it was simulating on a sub-atomic level in order to contain all the information contained within that simulated universe on a sub-atomic level.

That's just the memory, it has to be that big. Can't store 2 bits of information within a single bit. Can't store 2 sub-atomic particles of information within a single one. Not even talking about the rest of the architecture required for such a computer. Just the memory.

If our universe is a simulation then it's stored on something that's larger than our universe is.

and how do you know sub atomic level is same size in the theorical real universe? as posibility imagine if our atom would be few million times bigger than the atoms that the real computer is build upon in that case it would be much smaller in real case thats why its pointless to start trying disprove simulation theory since all the rules our virtuated universe would use. could be made up

 

[Vigormortis]

snip

Much of Mr. Gates claims fall directly into a false equivalency fallacy.

"This thing over here, that isn't at all that other thing over there, looks kind of like that other thing if I fiddle with the graphical representation of it enough. Therefore, they are both one in the same."

This is the essence of his claim. He is claiming that there is computer code embeded in the very fabric of the universe because "super strings" resemble the structure of our computer codes, rather than our computer codes resembling the structures of "super strings" because our computers, like the rest of the universe, are constrained to the structures and "laws" of the sub-atomic.

Even he admits his idea is extremely "theoretical" (by which I believe he meant hypothetical[footnote]Which is still an incorrect description as, based on the principal of his claim, it can't even be tested.[/footnote]) So any attempt to use his ideas as proof is pointless.

Isn't it more logical to think that our computers mimic the structure and form of our natural universe than it is to think that our universe mimics our computer systems?

 

[Hagi]

and how do you know sub atomic level is same size in the theorical real universe? as posibility imagine if our atom would be few million times bigger than the atoms that the real computer is build upon in that case it would be much smaller in real case thats why its pointless to start trying disprove simulation theory since all the rules our virtuated universe would use. could be made up

I don't.

But it adds yet another assumption to the simulation theory. Not only are we to assume that our universe is simulated we're also to assume that that other universe is under entirely different physical laws in order to enable that simulation.

As I've said, for I believe the fourth time now, I'm not trying to disprove this theory. I'm trying to cast doubts on it.

 

[Vigormortis]

and how do you know sub atomic level is same size in the theorical real universe? as posibility imagine if our atom would be few million times bigger than the atoms that the real computer is build upon in that case it would be much smaller in real case thats why its pointless to start trying disprove simulation theory since all the rules our virtuated universe would use. could be made up

I think you missed the point of his post entirely.

It doesn't matter how "big" the "atoms" are. It's about information.

Let's say you were to try to simulate the effect of two golf balls spinning in two different directions. But your simulation involves a single bowling ball spinning in one direction.

No matter what you do you can never simulate the same effect of the two spinning golf balls with your single bowling ball.

Another point ,let's say assume that this other, mythical, "real" universe that our "simulated" universe exists in does have a completely different set of physical "laws" and properties. That would only complicate the simulation process even further as the simulation would need to approximate an entirely different set of physical laws while simultaneously simulating each individual particle within the virtual universe.

No matter what physics you're using, such a feat is ludicrous.

Quite frankly, the notion of a "simulated universe" is small-scale. It's almost entirely predicated on the idea that the universe is smaller than it actually is.

 

[drakonz]

and how do you know sub atomic level is same size in the theorical real universe? as posibility imagine if our atom would be few million times bigger than the atoms that the real computer is build upon in that case it would be much smaller in real case thats why its pointless to start trying disprove simulation theory since all the rules our virtuated universe would use. could be made up

I think you missed the point of his post entirely.

It doesn't matter how "big" the "atoms" are. It's about information.

Let's say you were to try to simulate the effect of two golf balls spinning in two different directions. But your simulation involves a single bowling ball spinning in one direction.

No matter what you do you can never simulate the same effect of the two spinning golf balls with your single bowling ball.

Another point ,let's say assume that this other, mythical, "real" universe that our "simulated" universe exists in does have a completely different set of physical "laws" and properties. That would only complicate the simulation process even further as the simulation would need to approximate an entirely different set of physical laws while simultaneously simulating each individual particle within the virtual universe.

No matter what physics you're using, such a feat is ludicrous.

if you are able to make up any physics and laws for universe everything is posible so using any logic we have is pointless. for example every particle being able to develop smaller particles to contain more information would make it posible for single ''atom''(if you can call it that) to produce enought space for virtual universe. the whole idea is quite interesting if you think about it, but it will always be imposible to prove or disprove so it will just stay in same idea level as religions (note i dont personaly believe in this idea but i like playing around with the idea of it)

 

[Vigormortis]

if you are able to make up any physics and laws for universe everything is posible so using any logic we have is pointless. for example every particle being able to develop smaller particles to contain more information would make it posible for single ''atom''(if you can call it that) to produce enought space for virtual universe. the whole idea is quite interesting if you think about it, but it will always be imposible to prove or disprove so it will just stay in same level as religions

Well, I would argue that most religions are continually being disproven as we learn more and more about our universe. And if the "simulation theory" is in the same realm as most religions; i.e. mysticism; than it's likely to befall the same fate. (as all mysticism should)

Not that "simulation theory" isn't an intriguing concept. I've been intrigued by the concept for years. I've read up on a wealth of literature on the topic. It certainly makes for interesting bits of science fiction. However, the more we understand of our universe the less likely it becomes. In fact, the more ludicrous it becomes.

And really, you've demonstrated yourself the reliance on assumptions the "theory" has. Saying, "if you are able to make up any physics and laws for universe everything is posible" is the core idea the entire "theory" is predicated on. It's the notion that 'if we assume this, this, and this, then it's possible that our universe could be a simulation'.

In many ways, it's simply bad science. Great philosophical concept, and very useful as an allegory for things within science, but not good science.

[sub]I kept putting theory in quotes as "simulation theory" is NOT a scientific theory. It's not even a hypothesis yet.[/sub]

 

[Strazdas]

It doesn't matter how fast the computer is. It's not about speed. It doesn't matter what computational system it uses. It's not about calculations.

It's about information.

The data of sub-atomic particles would have to be stored, all of it's information. Everything. The only thing that can contain all the information that makes up a sub-atomic particle is a sub-atomic particle or something bigger.

The memory of your super-computer, regardless of what processor it used, regardless of how that processor did it's calculations, regardless of how fast that processor was, would have to be at least as large the universe it was simulating on a sub-atomic level in order to contain all the information contained within that simulated universe on a sub-atomic level.

That's just the memory, it has to be that big. Can't store 2 bits of information within a single bit. Can't store 2 sub-atomic particles of information within a single one. Not even talking about the rest of the architecture required for such a computer. Just the memory.

If our universe is a simulation then it's stored on something that's larger than our universe is.

two words: file compression.
you know, the algorythms we use to store 5MB of data in 1 MB file? yeah, because repeating data can be compressed. and for example all hydrogen molecules are hydrogen molecules, so they are repeating data. thus the actual required processing memory can be far less than whole universe because there is a lot of repeated data in the universe. we just need to figure out how to process it by decompressing it only in parts, which means we end up relying on processing speed and timegaps (not real time simulation but frames of once a year or so on or it would take longer than actual universe).

im not saying his theory is right or wrong, merely that we can store information in lower amounts than the actual information.

as far as the theory itself, who is to say that the unvierse that simulates it isnt far larger than the one its simulating? or that simulation isnt just aproximating unimportant parts (such as anything outside of solar system since we never got out and never got to actually check it). you dont need to simulate whole universe to make whole universe look simulated from a single point of view in a very short period of time.

 

[Hagi]

two words: file compression.
you know, the algorythms we use to store 5MB of data in 1 MB file? yeah, because repeating data can be compressed. and for example all hydrogen molecules are hydrogen molecules, so they are repeating data. thus the actual required processing memory can be far less than whole universe because there is a lot of repeated data in the universe. we just need to figure out how to process it by decompressing it only in parts, which means we end up relying on processing speed and timegaps (not real time simulation but frames of once a year or so on or it would take longer than actual universe).

im not saying his theory is right or wrong, merely that we can store information in lower amounts than the actual information.

as far as the theory itself, who is to say that the unvierse that simulates it isnt far larger than the one its simulating? or that simulation isnt just aproximating unimportant parts (such as anything outside of solar system since we never got out and never got to actually check it). you dont need to simulate whole universe to make whole universe look simulated from a single point of view in a very short period of time.

Except nothing on a sub-atomic level is an exact copy of something else, everything has unique positions, slightly different properties, a slightly different temperature etc.

And the only way to compress that is by losing data, which would mean it's no longer a simulation on a sub-atomic scale.

As for your second point, I'm not going to repeat again what I'm trying to do, it's above bolded and underlined.

 

[Strazdas]

Except nothing on a sub-atomic level is an exact copy of something else, everything has unique positions, slightly different properties, a slightly different temperature etc.

And the only way to compress that is by losing data, which would mean it's no longer a simulation on a sub-atomic scale.

As for your second point, I'm not going to repeat again what I'm trying to do, it's above bolded and underlined.

Is it. can we calculate it? or could it just be approximation with randomization upon inspection? how does inspecting itself changes subatomic particles? thats a singular force in itself that would stnadardize some parts. not everything is that unique on subatomic level that we can spectate, let alone one we cannot/havent.

Yes, i understand what you are doing, i was just giving another angle on the theory that the guy appears to not have expressed. by no means i would say its proof or anything. i personally do not believe it, does not mean i cannot entertain the thought and see if i can put it in some kind of logic.

 

[Hagi]

Is it. can we calculate it? or could it just be approximation with randomization upon inspection? how does inspecting itself changes subatomic particles? thats a singular force in itself that would stnadardize some parts. not everything is that unique on subatomic level that we can spectate, let alone one we cannot/havent.

Yes, i understand what you are doing, i was just giving another angle on the theory that the guy appears to not have expressed. by no means i would say its proof or anything. i personally do not believe it, does not mean i cannot entertain the thought and see if i can put it in some kind of logic.

Let me take another angle.

I'll explain my point by, right now, actually building a complete and perfect simulation of our entire universe in the most efficient way possible.

As I've said the best and most efficient way to store a sub-atomic particle's worth of information is a sub-atomic particle. Which means that in order to contain the particles, and all their possible configurations, we need a space as large as our universe to contain them.

Which means I'm in luck because right now there's a space around me exactly as large as our universe.

Next we need sub-atomic particles, enough of them to contain all the information in our universe. Which, again, means I'm very much in luck because in that space I've selected there are enough sub-atomic particles to suit my need. Exactly enough, not a single one too many or too few. My lucky day!

Now, we need to load those sub-atomic particles with all the required information. Which means my luck continues because all my sub-atomic particles already happen to contain the exact information they need to contain.

Lastly, to do our processing on the interactions of these particles, we need to move and change them as appropriate to do as they would do in our universe. Luckily the space I've selected seems to have natural laws that do exactly that.

So, there we have it, a perfect and complete simulation of our entire universe in the most efficient way possible.

What does it look like? Our entire universe.

 

[Strazdas]

Let me take another angle.

I'll explain my point by, right now, actually building a complete and perfect simulation of our entire universe in the most efficient way possible.

As I've said the best and most efficient way to store a sub-atomic particle's worth of information is a sub-atomic particle. Which means that in order to contain the particles, and all their possible configurations, we need a space as large as our universe to contain them.

Which means I'm in luck because right now there's a space around me exactly as large as our universe.

Next we need sub-atomic particles, enough of them to contain all the information in our universe. Which, again, means I'm very much in luck because in that space I've selected there are enough sub-atomic particles to suit my need. Exactly enough, not a single one too many or too few. My lucky day!

Now, we need to load those sub-atomic particles with all the required information. Which means my luck continues because all my sub-atomic particles already happen to contain the exact information they need to contain.

Lastly, to do our processing on the interactions of these particles, we need to move and change them as appropriate to do as they would do in our universe. Luckily the space I've selected seems to have natural laws that do exactly that.

So, there we have it, a perfect and complete simulation of our entire universe in the most efficient way possible.

What does it look like? Our entire universe.

this is a solution but it is not the only solution. you are assuming we require subatomic simulation at planck time intervals. this is not necessarily true.

 

[Racecarlock]

What about a simplified simulation of the universe?

You know, like how sim city doesn't require an entire actual city to run on?