Slashdot Mirror
NASA Achieves Breakthrough Black Hole Simulation
DoctorBit writes "NASA scientists have achieved a breakthrough in simulating the merging of two same-size non-spinning black holes based on a new translation of Einstein's general relativity equations. The scientists accomplished the feat by using some brand-new tensor calculus translations on the Linux-running, 10,240 Itanium processor SGI Altix Columbia supercomputer. These are reportedly the largest astrophysical calculations ever performed on a NASA supercomputer. According to NASA's Chief Scientist, "Now when we observe a black hole merger with LIGO or LISA, we can test Einstein's theory and see whether or not he was right.""
From the article: "when two massive black holes merge, all of space jiggles like a bowl of Jell-O"
Wouldn't Kraft Foods have prior art on this?
....a machine that can tell me where my lost left socks have gone!
"Itanium chips actually get used"
The catastrophic results of merging Microsoft and Linux?
The hilarious results of merging Intel and AMD.
The unexpected results of merging a spinning Steve Jobs (Intel is Evil/Intel is the best, brightest, future of Apple) and the O'Reilly No-Spin Zone.
Those I'd buy tickets for.
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
What kind of framerate do you get on that machine when playing Half-Life 2?
...would he have developed General Relativity sooner, or just played WarCraft?
--- Attorneys Assisting Citizen-Soldiers & Families -
Based on observations, what percentage of black holes are non-spinning vs spinning?
Now when we observe a black hole merger with LIGO or LISA, we can test Einstein's theory and see whether or not he was right.
And if he's wrong then all the scientists can make "loser" signs at him on their foreheads...
Summation 2
anyone know if google has a science-nerd-jargon translator?
"Simplify, simplify, simplify!" Thoreau
And even more likely: Whether or not the computers performed the calculations correctly (the chips are made from Intel, and we all know the history of Intel screwing up floating point math)
Non spinning black holes?
Is there such a thing?
Democrats or Republicans. They are both taking us to the same place and they are not afraid of us anymore.
Not mentioned in the article of course, is that shortly after the simulation, the software collapsed in on itself as it underwent a Massive Total Existence Failure.
There are experimenters. The guys who ran the simulation were experimenters.
There are theoreticians. Einstein was a theoretician. He asked relatively simple questions and followed the logical consequences. I suspect that having to use a computer would have been a giant distraction and might have delayed or prevented the theory of relativity.
Watching massive things merging.. jiggling like jell-o... Good heavens, space is a pervert!!!
Slashdot Burying Stories About Slashdot Media Owned
They're not black holes, they're just a result of Intelligent Darkness. Seriously, why do we teach kids today the theory of black holes without not also teaching them about Intelligent Darkness?
I don't know about you, but I already give enough of my money to publicans on a Friday night...
"I Know You Are But What Am I?"
Scientists has been doing similar calculations for a long time. For example
Larry Smarr, "Gravitational Radiation from Distant Encounters and Head-On Collisions of Black Holes: The Zero Frequency Limit," Phys. Rev., D15, 2069-2077, 1977.
I cite this paper because Larry Smarr is one of the Nasa panelists for this project, and I heard his talk on this paper at the University of Texas at Austin in the late 1970s. Come to think of it, I remember seeing one of the other panelists, Joan Centrella, at the same talk.
OK, I'm no general relativist, but I am a computational physicist -- what could the article possibly mean when it says earlier attempts were "plagued by computer crashes -- the equations were far too complex"?
I can imagine a situation where a poorly-arranged computation of an equation might give you an underflow in an intermediate result, or where a badly-arranged summation might give you noise. But crashing the computer? Sounds more like array-bounds, which can happen no matter how simple the equations are.
2*3*3*3*3*11*251
What is useless now will someday be useful.
Exempli gratis (and it's way out there):
Using this new data, someone observes a black hole merger. It doesn't fit the data. Relativity is redone, so to speak. Someone sees a great way to unify Relativity and quantum mechanics because of the new formulation. Bam. Like that, unified theory of everything. Those spinning superconductors generating magnetogravitic fields are understood. Artificial gravity and anti-gravity are discovered. Moon-flights are near cheap after a while. Etc. etc.
Saying "I don't see any results coming out of this tomorrow so this research is useless" is about as shortsighted as one can get. It's akin to foreign aid: sure, it gets us little immediate benefits, but the long-term stuff can really pile up.
What is the actual outcome from this research?
more knowledge about the universe and how it might work.
Will this help create more energy-efficiency in the world?
maybe, who can say what future developments and understanding of this area of physics will bring.
Will it help us find technology that humanity can actually use to make a better society?
maybe, see above. it depends on the definition of "better".
when general relativity was first thought of in 1915 there was no application, for the average person. today GPS relies on general relativity.
Will it increase our safety, or decrease power of madmen and dictators?
the obvious answer is probably not. and while these are important questions, this one is not topical in this discussion.
--meh--
How about making science progress by testing a part of one of the most important theory in physics? It's not my funding, however I'd love my country to invest more in science even if only for the sake of science. We're in an era where everything has to be justified by money, it feels like the Dark Age of information. I'm waiting for the next era where new thoughts, science and knowledge progress get some value back.
Call me utopist if you want, but finding something that "increase our safety, or decrease power of madmen and dictators" gets the #1 naive award (always thinking big shields and weapons, what a world).
I realize that this doesn't fit nicely into your libertarian view, but we often do science just for the sake of doing it. Knowledge in and of itself is a good thing, and funding some cycles on a computer that would otherwise be simulating nukes or finding prime numbers doesn't seem wasteful to me at all.
So if this experiment shows us that Einstein was right about gravitational waves, and those waves can tell us so much about the universe, I wouldn't call it a waste of money. Of course now we have to go through the trouble of actually detecting the bastards...
And I whole heartily encourage all patent and IP lawyers to go to those black holes and ether Subpoena them or deliver a notice of possible infringement.
This should solve all lot of problem here on earth as well, if we can get them to all go.
Unless that is the Black hole decides to show up for its court date.
"You apparently use a supercomputer to generate a problem to a hole that sucks everything in."
I really think the goatse guy could claim prior art on this.
What if the Hokey Pokey really is what it's all about?
HL2 is singlethreaded so the performance would be the same as on one Itanium. Also x86 code has to be emulated on Itaniums = slow. Oh and no GPU which means pixel/vertex shaders would have to run on software. Educated guess: 0.1 fps.
That's nothing--the WB and UPN are merging in September, producing a vortex of TV so sucky that not even brain cells will be able to escape.
The interesting question is whether the CW black hole will rotate or not. I for one hope that TV execs will be able to sit on it and spin.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
This isn't the only testing that NASA is doing of Einstein's theories. For those that are interested, there is also the Gravity Probe-B. Really interesting stuff!
In case anyone was wondering how Columbia stacks against their rig, check out:
http://www.top500.org/
Here's the November 2005 list:
http://www.top500.org/lists/2005/11/TOP10_Nov2005. pdf
It shows Columbia with:
51.87 Rmax (teraflops/second).. It also states that it moved from #3 ranking to #4.
What, they sent another budget request to Congress ? :-)
You can't do meaningful experiments without some idea of what the theory says will happen. Numerics of this sort provide that for complex physical cases which are essentially impossible to work out with pen and paper. So yes, this is a step towards getting knowledge of the universe and how it might work.
Also, understanding does NOT require the tie to experiment since you can have mathematical understanding of a particular theory independant of whether that theory properly models reality. For instance, I can go and work out what orbits look like in a five dimensional space. If I go and check my results versus reality 'hey, it doesn't match up!'. So from that point of view, all I've learned is that space on that scale isn't five dimensional. But lets say I don't even bother to check versus reality. I've still learned something about the mathematical properties of the theory and I've gained intuition about how things behave - namely, I've learned that closed (classical) orbits seem to only be able to exist in 3d. How strange! And I've learned that if I had a situation where the system was, say, restricted to a lower dimension (examples in electromagnetism) then I can expect large changes to the dynamics. Or perhaps a better example is, we can learn a lot about phase transitions in three dimensions by doing problems in four dimensions where they can be solved exactly and then doing an expansion around the four dimensional solution to approximate the solution in 3d. That approach doesn't depend on the underlying Hamiltonian you're solving being the correct one for some physical system - it is a purely mathematical understanding which can generically be applied to many different theories. So the benefit is, in the future when I find a better Hamiltonian for my phase transition, a better dynamical theory for gravitation, etc, I can apply the techniques I've learned from before to those as well.
You use his theories to construct and run a model, and then you compare the results of that model to what you can observe in the sky. The differences between what is observable and what the model indicates are where the new knowledge is, even if things don't match up.
- Waterworld - $175,000,000
- Lethal Weapon 4 - $140,000,000
- Dante's Peak - $116,000,000
- Star Wars I: Phantom Menace - $110,000,000
(Those are production costs.)The key here is not really what the model looks like. It's how the model compares to real life. If when LIGO comes online, they detect waves that match what the model predicts they should detect, that gives experimental support that the equations the model is based on are correct. Also, in your example the equation is part of the theory, which is that 7=13, so then the model is 7*2 and the result is 26. If you do an experiment with counting blocks and combine two groups of 7 blocks yet find yourself with 14 blocks instead of 26, you should conclude that your theory is incorrect.
This analogy is kind of clumsy because you're essentially attempting to redefine the meaning of numbers, then directly compare the results of operations under the original number definitions. I think an equivalent situation would be to define an orange as an apple, then ask why, when I show you both an apple and an orange (according to my definitions), the two objects in my hand are different.
A better example is to theorize that gravity is proportional to mass (g = G*m1*m2/r^2). You can build a model based on this equation where the gravitational attraction between two masses at a certain distance with the known gravitational constant works out to be 2 Newtons. Then you can actually get the two masses, hold them r meters apart, and measure the force required to keep them apart. If it's 2 N, your theory looks good (it's not technically proven, but it's one step closer). If you get, say 1.9 Newtons, there are several possibilities: you theory is wrong (in this case we know it's right as far as classical physics is concerned), you made the model wrong (ie, you suck at math), or there were other factors influencing the experiment that you failed to account for (perhaps friction in your scale).
Won't building a model based on an equation automatically prove a theory that is based on that equation?
No. In Physics a theory makes claims that can be falsified by an experiment. The theory (general relativity) is already there and the experiments will be carried out by LIGO and LISA (the latter having been delayed indefinitely thanks to Bush's plans).
However, we strongly assume that General Relativity must break down at some point and give way to some theory of quantum gravity. There are several such theories and we simply don't know which is correct, if any. So if one of these experiments showed a deviation from general relativity that would be very exciting.
OS Reviews: Free and Open Source Software
I'm a recent member of this group, so I'd like to put in my 2 cents.
:) (some competing groups use Cactus which is C++ based, although it also allows C and Fortran).
1) This is a first -- no other group has achieved this before. yay! (after decades of work!)
2) This is hard for the following reasons:
a) since you are doing calculations near (or on/in) a black hole, you tend to get a lot of
infinities, which 1) crash your code and 2) exacerbate your errors
b) for most simulations, your grid remains fixed. For black holes though, they *deform* the
spacetime around them -- which means your grid points have to move (in a non-predictable
manner)!
c) what happens when two black holes merge is not well understood (ie, what should happen?),
so this is new science
d) initial data is hard to get and unreliable. If two black holes are far apart, you can
write an exact solution (at least within some error), but to get them close to where they
are interating, you pretty much need this kind of simulation anyways. This is such a large
problem that there are only a handful (a dozen or two?) initial data sets currently.
3) Everything is written in Fortran!
4) It runs on a variety of architectures (x86, Itanium, PA-RISC, Alpha, etc etc)...pretty much
anything that supports ifc (faster) or gcc.
5) There are several approaches to some of the issues above, from puncture splitting (using a
different spacetime metric like 1/r vs r to remove the singularity), excision (not evolving
inside the event horizon, since that's not "interesting" anyways), and other methods. Our
new method actually doesn't need any of those "tricks", which is pretty interesting.
6) This data helps drive the LISA and LIGO projects from a theoretical standpoint--basically
knowing what kind of gravitional waves they should be seeing, and to correlate what they see
and what their data may represent (ie, if you see a waveform like this, this means that it's
two merging black holes, vs just co-rotating black holes).
6a) We study black holes b/c they are pretty much the only thing that'll generate detectable
gravitational waves.
so yay!
"You have the option of insanity. I do not. And that makes me crazy!" - Brian to Angela, My So-Called Life