This visualization of dark matter is one-thousandth of the gigantic Bolshoi cosmological simulation, zooming in on a region centered on the dark matter halo of a very large cluster of galaxies. (Credit: Chris Henze, NASA ARC)
If you're going to create a virtual universe, you're going to need a big computer — like the Pleiades supercomputer at NASA's Ames Research Center in California's Silicon Valley. Researchers have just made the most accurate computer simulation showing the evolution of large-scale structure in the universe, known as the Bolshoi simulation, available to astrophysicists around the world.
Bolshoi (which takes its name from the Russian word for "grand" or "big") took in data from ground-based and space-based instruments, including the best readings of the big bang's afterglow from NASA's Wilkinson Microwave Anisotropy Probe, or WMAP. Then it used 6 million CPU hours on Pleiades, ranked as the world's seventh-fastest supercomputer, to crunch all that data into a virtual representation of the universe evolving over time.
The time-lapse simulation occupies nearly 90 trillion bytes of memory, or the equivalent of nearly 10,000 typical movie DVDs.
The first two papers in a series describing the simulation have been accepted for publication in The Astrophysical Journal. "A lot more papers are on the way," one of the co-authors, physicist Joel Primack, said in a news release from the University of California at Santa Cruz.
So far, the simulation has been in close agreement with what astronomers are seeing in the actual universe.
"In one sense, you might think the initial results are a little boring, because they basically show that our standard cosmological model works," Primack said. "What's exciting is that we now have this highly accurate simulation that will provide the basis for lots of important new studies in the months and years to come."
The standard model suggests that only 4 percent of the universe's mass-energy content consists of ordinary matter — the kind that we can see. Another 22 percent is cold dark matter, which can be detected only by its gravitational influence. Physicists surmise that dark matter is made up of exotic particles that interact only weakly with ordinary matter, but they haven't yet identified any of those particles. It's the weightiness of dark matter that is thought to shape galaxy clusters into a "cosmic web," which you can easily see forming in the animation above. (Remember to go full-screen and HD for optimal effect, or check out this music-enhanced Vimeo version.)
The biggest constituent of the cosmos, at least based on current models, is dark energy: This mysterious energy, which is thought to account for around 74 percent of cosmic density, serves to counteract the force of gravity and cause the accelerating expansion of the universe. Its existence is required to reconcile cosmological theories with WMAP's observations as well as observations of distant supernovae — but no one has figured out what it is, which has led some astronomers to look for alternative theories.
Primack, who directs the University of California High-Performance Astrocomputing Center, said a close analysis of the Bolshoi simulation could help point the way to better explanations for the dark-energy effect.
"These huge cosmological simulations are essential for interpreting the results of ongoing astronomical observations and for planning the new large surveys of the universe that are expected to help determine the nature of the mysterious dark energy," he said.
The first paper based on Bolshoi analysis focuses on the role of dark-matter halos in the universe's development, while the second paper looks at Bolshoi's predictions for the abundance and properties of galaxies. The researchers have found that the simulation correctly predicts the number of galaxies as bright as our own Milky Way that have satellite galaxies as bright as the Milky Way's major satellite galaxies, the Large and Small Magellanic Clouds.
But this is just the tip of the iceberg: So far, less than 1 percent of the Bolshoi project's output has been released, Primack said. The Bolshoi simulation computes the evolution of a cubic volume measuring about a billion light-years on a side, following the interactions of 8.6 billion particles of dark matter. A variant of the simulation, called BigBolshoi or MultiDark, was run with the same number of particles in a volume 64 times larger. Another variant called MiniBolshoi is currently being run on Pleiades. It focuses on a smaller portion of the universe with higher resolution.
This all sounds pretty deep, but fortunately, the Bolshoi team has produced plenty of beautiful videos and illustrations that will delight even those who can't tell a baryon from a meson. For still more background about Bolshoi, check out the news releases from New Mexico State University, Ames Research Center and the High-Performance Astrocomputing Center.
Update for 5:50 p.m. Oct. 7: In a follow-up phone call, Primack told me that "the agreement between predictions that come from the simulations and the actual observations are really getting spectacular." The previous top-of-the-line virtual universe, known as the Millennium Simulation, showed galaxies as being "much more clustered than they actually are," he said, while the Bolshoi version is "bang-on." Primack said still more revelations are coming from the Bolshoi team. "It's like things are coming into sharp focus," he said.
More about cosmology:
- 3-D map looks at universe of 11 billion years ago
- 'Accelerating universe' could be just an illusion
- Hurry! Only a trillion years left to study the big bang
- Interactive: Beyond the Big Bang
Authors of "Halos and Galaxies in the Standard Cosmological Model: Results From the Bolshoi Simulation" include Anatoly Klypin, Sebastian Trujillo-Gomez and Joel Primack. Authors of "Galaxies in LCDM With Halo Abundance Matching: Luminosity-Velocity Relation, Baryonic Mass-Velocity Relation, Velocity Function and Clustering" include Trujillo-Gomez, Klypin, Primack and Aaron J. Romanowsky.
Connect with the Cosmic Log community by "liking" the log's Facebook page, following @b0yle on Twitter or adding me to your Google+ circle. You can also check out "The Case for Pluto," my book about the controversial dwarf planet and the search for other worlds.


This is so funny. They take a computer and make a program that acts the way they THINK things happen and then pat themselves on the back. If it doesn't work the way they THINK it should they modify it until it does. Yeah, it's definately a simulation of the universe. They way they THINK it happened.
Did you stop to think that maybe they can use the model of the universe the way they think it is to find and correct the flaws in the way they understand the universe?
Or, are you really just against science all together because it does not offer the easy answers you are looking for.
I don't think you understand how the scientific process works...
The neutrino was found using a similar process to this.... but then experiments were actually designed to detect them in the lab.
Until dark matter is detected, it is just a way of sweeping the cosmological dirt under the carpet.
Jack...
You obviously don't understand how this works.
The model they created takes a set of assumptions about the origins of the universe, and then runs them through a time-lapse based on observed physical principles to determine what the universe should look like. If their assumptions are wrong, the model would produce a a wildly different looking universe than the one we have. That fact that it didn't means that we are on the right track.
This is a Confirmation of Theory, not a redundant restatement of terms.
Jack, read the article:
WMG-21, The essential problem here is that we know practically nothing about exotic dark matter, including whether it exists.
What you end up doing is just ascribing to exotic dark matter whatever properties are necessary to get a good fit to our Universe. Not really very good science, especially since there are many other first-principles theories that need to be investigated (or exotic dark matter needs to be detected and experimented on).
I hesitate to believe such simulations because of the extraordinary failure to compute an accurate model of planet formation, a much less computationally intensive task.
Now this is really cool stuff. Every time I see this video I think of the possibility of universal intelligence. Are we living inside "God"?
On another subject, this is a very impressive computer, so why can't they create a neural network to predict earthquakes?
On the first point, the analogy of God the Author and the author as god reveals a perspective where 'living inside God' could be viewed in another way entirely: not like your blood cells live inside you, but the way your imagined things 'live' inside your imagination.
On the second, because the mechanics of earthquakes are far less understood. Stress propagation is a far more complex field of physics than even relativistic gravity.
First, I am not interested in an imaginary God, I am looking for the real God. I have no faith in "man's" religions. I trust Nature far more than man.
Second, if we can input enough real time seismic signals, the effects of major earthquakes, into a neural network, can the network develop itself to figure out the cause, the time of major earthquakes.?
your tax dollars at work
I know! Isn't it awesome!? ^^
I wish they'd spend more of my tax dollars on science like this than they do on war.
Если бы это не было для наших сильных вооруженных сил, то мы говорили бы на русском языке теперь. :-)
Если бы это не было для наших сильных вооруженных сил, то мы говорили бы на русском языке теперь.
Oder Deutsch sprechen.
Translations: If it weren't for our strong armed forces, we'd be speaking Russian now ... or German.
Alan Boyle - 言語学者、学者、非常にハンサム!
Ich spreche Deutsch! :)
My tax dollars at work, and I love it!!!
Mr Eous, I would rather have my taxes spent like this than on more corporate bailouts and more wars.
dark matter in the string theory... Do gravitational effects traverse the multiverse?
I tend to agree...dark matter in string theory???
I may have missed something here or maybe their code is a little top heavy. It was pretty and all, but what can it predict? You'd think I could find out the winner of the Dallas Detroit game this Sunday for all those CPU cycles. If it can't give me that, I'm supposed to believe that it tells me about the universe? Please. I know another BS word besides Bolshoi. How about the Bears-Carolina? At least tell me who it likes.
What? Are you stupid or something? Are you one of those brainless beer belly sport belching jerks who tease nerds because they intimidate your intellect? You may lack intelligence but at least you can hit stuff and hurt things, huh? Now go play ball with the other children, scientist are trying to work here.
I don't know much about football but you'd have to be an idiot to trust that bum-@!$%# Cutler to come through with anything.
As a novice 3D modeler I can respect this effort with awe. And what we see above is 1/1000 of the whole simulation, wow! How I would love to get hold of the algorithms used for the particle interactions. What would be the properties and methods of dark matter? Can't wait for the MiniBolshoi to come out.
you said it Clint...I find that I dont need Hilbert level math to somewhat figure out a field/interactions when presented in scientific 3D simulation.
Clint I still have my Toaster4000/Lightwave setup (working).
Who won last nights game?
at last...a place for my virtual ego...
Can't be correct, the Universe is expanding at an ever accelerating rate and this model shows huge groupings of matter which would create gravity problems. The center of the Universe cannot be full of matter. and the edges cannot have huge conglomerations. It doesn't show anything but a waste of puter time with faulty logic.
papa...as I understand various loop theorys the expansion may only appear as such due to our relative position in this place. The red shift may be only partially understood and our historical analyslis poorly derived...
"a very tiny part of the Whole"?... compared to an atom we're quite large, possible somewhere in the middle in relative perspective to the whole .
The Universe is simply expanding at an ever excellerating rate with the furthest Galaxies red shifting away fueled by Black Mater/energy toward the other Universes, Billions of them, many of which have already combined to create new universes ever expanding, from a vast central organization of chaos.
Simply put, our's is but a very tiny part of the Whole.
Can someone tell me what a computer hour is? Is that like dog hours? Because 6000000 hours is like 600+ years. They could not have finished the simulation yet.
the computer can assimilate, correlate and spit out data at a significantly faster rate then our brains. So yes, in comparisom, its like dog hours.