Simulating Supernovae with Graphics Cards

astroboy writes "As graphics cards get more powerful, Los Alamos and Utah scientists have developed a package, Scout, to use those usually-languishing FLOPs to do simulations, and to visualize of them on the on the run. As an example, they have released movie of part of the evolution of a core-collapse supernovae"

Comments

Can you imagine a Beowulf cluster of these?

[TripMaster Monkey]

From TFA:

To make the technology much more powerful, McCormick is working on a version of Scout that will work when several computers are linked together.

I guess they can.

^_^

They can do this....

[Bloodlent]

But they can't bring back Suprnova? Dammit! How am I gonna get my Desperate Housewives?!

Ulterior motive?

[TripMaster Monkey]

From TFA:

Peter Schröder, a computer simulation expert at the California Institute of Technology, believes graphics processors have great potential for scientific research. "There is a real market driving this hardware that we can use for scientific computation," he told New Scientist.

Actually, Peter and his buds just got sick of getting scragged in DeathMatch because the video cards in their lab computers are teh SUXX0R.
Now, they have a blank check to get whatever video cards they want.

^_^

"Captain, the servers, they cannae take no more!"

[Chris Tucker]

Well, you've destroyed those nice scientists computers.

Go to your rooms and I want you to think long and hard about what you've done!

Been there done that

[a_greer2005]

I have seen graphics cards go supernova - just overclock one and you can see it too...FOR REAL, no 3D sim crap,..

Re:Been there done that

[HazE_nMe]

I tried... When I OC too high it just locks the computer up and scrambles the onscreen fonts... For a real show turn off the lights, voltmod a card, remove the hsf during HL2, and pour saltwater on it when it starts smoking. ;)

I myself have done similar things on Linux

[Winckle]

boot up the GIMP: filters>light effects>supernova dunno what the big deal is?

I'd have thought it obvious.

[jd]

Now they've simulated supernovae on their graphics cards, they want to try creating one for real in their web server.

Movie torrent

[slavemowgli]

BitTorrent for the movie, in case of Slashdotting: here

torrents and mirrors

[xbmodder]

the PDF http://xbmodder.us/scout.pdf the torrent for the .mov: http://xbmodder.us/Scout.mov.torrent The torrent for avi (divx4) http://xbmodder.us/Scout.avi.torrent

Article Correction (Los Alamos)

[smoany]

Usually, I think that New Scientist is pretty accurate as far as laymen-science articles go, but they've let a big mistake slip be.

From the article:
"The Scout programming language, developed at Los Alamos National Laboratory (LANL) in California, US, lets scientists run complex calculations on a computer's graphics processing unit (GPU) instead of its central processing unit (CPU).

Los Alamos National Labs (LANL) is based in (fittingly) Los Alamos, New Mexico. it is currently operated by the University of California, which has contracted for the ability to manage the lab. This may have caused the confusion.

Also, Lawrence Livermore National Labs (LLNL) is based in Northern California, so that may have caused the confusion as well.

Not a terribly serious concern, but their fact's should be straight. The lab is not in California, it is in New Mexico... Editors: shame on you!

Oh, the irony

[smoany]

Of course, in my hurry to post my response, I let a few big editing slips pass by...

It should be "Slip by" not "Slip be"

Also, it should read "facts" not "fact's".

Oh well. I never said I was good at editing, only that New Scientist should have been.

MPAA Cease and desist

[Anonymous Coward]

MOTION PICTURE ASSOCIATION OF AMERICA, INC.
15503 VENTURA BOULEVARD
ENCINO, CALIFORNIA 91436
UNITED STATES
PHONE: (818) 728-8127
Email: MPAA23@pacbell.net
Anti-Piracy Operations

Date: June 11, 2005

Dear slavemowgli:

The Motion Picture Association of America is authorized to act on behalf of the following copyright owners:

Columbia Pictures Industries, Inc. Disney Enterprises, Inc. Metro-Goldwyn-Mayer Studios Inc. Paramount Pictures Corporation TriStar Pictures, Inc. Twentieth Century Fox Film Corporation United Artists Pictures, Inc. United Artists Corporation Universal City Studios, Inc. Warner Bros., a Division of Time Warner Entertainment Company, L.P.

We have knowledge that you posted a torrent to one of our client's movie (The Scout : http://www.imdb.com/title/tt0111094/) and are demanding that you withdraw this link at once.

Failure to do so will make you loose more then just your modpoints :P ...

Re:What's going on in the movie

[Toxygen]

The experiment isn't done to show us what supernovea look like, it's more like they've discovered that the gpu from their graphics cards are better suited to the types of operations required to define a supernova than a traditional cpu. The video is just a shiny bonus.

What about precision???

[geneing]

If I understand correctly graphics cards don't implement IEEE floating point standard. This means that you can expect all kinds of wierd problems with complicated floating point computations ahref=http://www.cs.berkeley.edu/~wkahan/ieee754st atus/754story.htmlhttp://www.cs.berkeley.edu/~wkah an/ieee754status/754story.html>. I wonder how they know they can trust results of their simulations.

Re:What about precision???

[mmp]

NVIDIA's GPUs are only one or two bits short of perfect 32 bit IEEE floats. (ATI's are still at 24 bit floats.)

See Karl Hillesland and Anselmo Lastra's cool work on measuring this error on current GPUs, GPU Floating-Point Paranoia for much more information.

-matt

Re:What about precision???

[non0score]

Err, what? All graphics cards that implement ARB_color_buffer_float has to implement IEEE 32-bit floats, as stipulated by ARB extension specification. (of course, this is assuming that the scientists are using the color buffer to encode information)
http://oss.sgi.com/projects/ogl-sample/registry/AR B/color_buffer_float.txt
Basically, any up-to-date ATi or NVidia gfx cards are capable of true IEEE 32-bit floating point numbers. What really worries me about the research is that they're not using 64-bit!

Re:What about precision???

[geneing]

It's not a question of the size of a fp number. There are many subtle points in designing "safe" floating point arithmetic.

IEEE 754 compliance makes fp operations slower, which is why hardware doesn't often support it (famous example Cray where SQRT(1-COS(X)) could return with an error root of a negative number).

Roundoff errors might not matter for graphics (who cares about being one pixel off?), but it is a huge problem for numerical computations.

Also, does GPU signal overflow/underflow/division by zero??

Re:What about precision???

[TrurlTheConstructor]

I did RFTA, and they don't actually run the simulation on the graphics card. On the contrary, they had to downsample the data from 320^3 to 256^3 just to fit it into the GPU's memory. All they did in the GPU was a bit of post-processing (and the rendering, which looks nice enough).

In a more general sense, I wouldn't "trust" the result of a hydro-only simulation of a SN explosion in any detail. Too much physics left out, and a lot of chaotic dynamics which are only barely resolved (or not at all). An experiment like this is indicative of the likely behavior of a general class of models, but it's not a prediction.

Los Alamos and supernovae collapse....

[mikael]

... Are they designing a nova bomb?

Re:I must say

[joepeg]

I believe this is the Mike Batt version.

Source? License?

[tbo]

I know this is slashdot, and I appreciate all the Beowulf cluster jokes, especially since they're actually appropriate here, but nobody is asking any meaningful questions. By my calculations, the noise-to-signal ratio is illegal div_by_zero.

Where can I get Scout? What is the license? What platforms are supported? I'm working on an open-source scientific computing package for doing quantum simulations, and I'd like to use Scout for visualization, but this article provides no information on where to get Scout or even if the licensing would allow me to use it.

It's also not clear exactly how you'd link Scout up with an existing app. Does Scout produce machine code that you stick into your app somehow? Are there C or C++ wrappers for using Scout?

Sad days.

[imsabbel]

Slashdot should rename itself to "news for computer kiddies and layed of cynical IT-veterans who lost touch with technology".

Both this story and the last one (the quad core one) were nice technical stuff, perfect for nerds.
And lets take a look here. at the time of that posting , only 2 or 3 comments are even remotely touching the subject. The rest is stupid jokes and dumb ranting.
The quad core article is even worse, were the only non-joke posters are to stupid to tell apart SMT and dualcore.

Also it seems to be a sad trend that the initial reaction to ANYTHING even slightly technical/scientific seems to be a self preservation (" im not stupid, this stuff is just ununderstandable !!!11") joke posting.

Nice to see this idea surface again

[Frumious Wombat]

Every few years it seems that some variant of using the GPU comes back for scientific computing. I seem to remember in the early 90s a group using the graphics card for the additional memory it could provide. I run quantum-chemistry simulations for a living (basically large quantities of matrix algebra), so anything that could speed up calculations currently taking weeks would be appreciated.

Personally, I'd like to see someone port BLAS (or the ATLAS variant) to a set of standard gpus, so that we could speed up matrix ops. I've been hoping for a more general-purpose solution making it to market, such as the old Celerity strap-on vector unit except for modern IA32/AMD64/PPC, but this may be the better solution.

For those of us who don't have a budget for a Power5 or Cray system, maybe a pair of PCI-e cards running the matrix algebra and FFT routines would be the way to go.

Re:Nice to see this idea surface again

[sadangel]

It's been found that GPUs, despite their impressive floating point capabilities, can't compare to heavily-optimized and cache coherent CPU implementations of large matrix operations, such as ATLAS. The exception is when the result is to be displayed anyway, as in scientific visualization and Scout. The real drawback of GPUs is the readback speeds. When the result is done, if it isn't to be displayed, it must be read back into the CPU memory. This is notoriously inefficient. PCIe is improving this, but it's still a serious problem.

Re:Neat stuff..

[sadangel]

No, that's a different project. As the only one from Utah directly working on this project, I can tell you it wasn't me. Utah has another project that is also quite interesting. csafe is involved in simulating explosions and fires. It was likely someone working on that.

seti@home

[frostilicus2]

In a similar vein, the seti@home project is currently developing a new project called "Astropulse" to scan the skies for optical signals from ET. This is also designed to use GPU code to perform the signal analysis. (It would be interesting to see how this woud perform on a PS3, especially now the PS3 is rumoured to ship with Linux pre-installed)