NASA | Climate in a Box

Climate modeling requires massive computational power. Until recently, that power required room sized machines with daunting technical and logistic requirements. But new advances in computer design, including hardware and software, continue to facilitate a paradigm shift. In an effort to broaden and democratize climate research tools, NASA has begun to facilitate the operation of new desktop sized supercomputers, with the goal of making it substantially easier for more researchers to do meaningful work on vital and essential questions for our world.

Want more? Subscribe to NASA on iTunes!

Or get tweeted by NASA:

[WilBlack]

NASA and other scientists use Fortran partly because its legacy code but also because it was the only thing that could run on across different super-computing architectures and memory configurations. Also you do not need a high level languages to do numeric simulations so there is no need to put the effort into rewriting old code. For the post-processing of the model output data (which is usually all the public sees) you can use whatever you want.

[intrlper]

I seem to remember you can mix C and Fortran quite easily; you'd need to write a few interface routines, but apart from that, there should be nothing stopping people from using Cuda with their Fortran programs.

[intrlper]

Correction. nVidia Tesla processors. Based on the Quadro designs, but without dedicated video hardware, they're pure parallell number crunching units.

[WilBlack]

I don't thinks it a problem. There is just no need to use high level languages for numeric computation on this scale.

[chinito]

hm the university of antwerp also developed a super computer (FASTRA II) for only 6000euro is it the same one?

[beanNronin]

gee, that was my first job programming. I worked up from there.

[intrlper]

The machine in question is basically a desktop computer with a bunch of nVidia Quadro GPUs thrown in.

[samipso]

But you don't need windows programs for science. Scientists make their own applications. I don't know what the odds are between windows and linux for science use, but i can imagine linux being popular.

Besides, i don't see the problem. Most programs have a counter part on other OS's. Windows just has more.

[senoctar]

hey, it's grandpa fortran 90
looks like NASA has quite some legacy code problems

[princeofexcess]

linux for the win... doesnt crash is open source and its free... i still own a windows though :( since most applications are already written for it. Windows core should be available so other companies systems could run windows programs but no it has to be a monopoly... grrrr... and errors are killing me

[princeofexcess]

its not only cheaper. Its safer its faster and its better. The only problem is no programs for it. Who needs an os if he cant use anything on it. :/
thats why MS is a monopoly because noone can make an OS that can run windows programs (especially games) so companies are forced to make programs for windows because they will get most buyers from that. And there is more programs for windows and the vicious cycle continues.

[LadyTink]

Good video... didn't know about this :D

[estefanello]

how about put this computer to run DNA, RNA and protein folding? It could be very usefull!

[MichaelAye]

oh, look, seems like senoctar has some 'need-to-reinvent-the-wheel' problems.. ;)

[philbert_io]

i still think its soo stupid that people are still using fortran even though f90 is the new standard etc, no one really uses it tbh!

[samipso]

Linux is the cheaper solution for applications like this. Open source is cheaper in every aspect. Which is why many governments around the world use it for most of their things.

Windows is just a ripoff and a IT disaster.

[beanNronin]

Except for people running supercomputer models that require gazillions of computations. Go back to your business apps, junior, and get off my lawn!