Researchers tout 20 million processor-strong supercomputer to study climate change

It looks like a group of researchers at UC Berkeley have come up with a rather unique way of solving the problem of getting supercomputers past the processing power / energy consumption barrier, with them now touting the possibility of using millions of low-power embedded microprocessors instead of conventional server processors. That tantalizing prospect has apparently already lead to a deal with Tensilica Inc, which will provide the Berkeley researchers with some of its Xtensa LX extensible processor cores to use as the "basic building blocks in a massively parallel system design." Ultimately, the researchers say they could one day build a massive supercomputer consisting of 20 million embedded microprocessors at a cost of $75 million, which they say would have a power consumption of less than 4 megawatts and a peak performance of 200 petaflops. That, they say, would be enough for it to create climate models at 1-kilometer scale or, as the researchers put it, more than 1,000 times more powerful than anything available today.

[Via TG Daily]

[Ike Turner @ May 7th 2008 12:32PM]

Researchers tout 20 million processor-strong supercomputer to study biology and aging diseases...and save lifes

[clak @ May 7th 2008 12:38PM]

Finally, a computer that will run Vista.

Pimps forever!

[Flashpoint @ May 7th 2008 12:41PM]

THIS IS AN ABSOLUTE WASTE OF TECHNOLOGY.

I could use this to run Crysis.

[stefan @ May 7th 2008 12:51PM]

this monster may have potential. But it will have one impact on global climate which i'm already certain about.

[Kiwi616 @ May 7th 2008 1:08PM]

Skynet here we come...

[Jason Rhyne @ May 7th 2008 2:40PM]

I heard about that computer. Its called the Playstation 4 or maybe PS5 depending when the next Crysis game comes out.

[Gorillamonk @ May 7th 2008 3:23PM]

is it really necessary?
I mean really?

I don't see how knowing more of what the climate is changing will do. Have you seen the hurricanes and tornadoes that we have at least a week of warning, that kill 20k people?

I think we need a better way of handling everything first. Then, we can worry about trying to tell what mother nature will do next.

[rob @ May 7th 2008 4:12PM]

Perhaps someone should point them at an IBM BlueGene which uses Power chips, which come from the embedded market side also.

Now admittedly the current version will not scale to 20 million (only a paltry 884,736 to get you at 3 PetaFlops), but future versions should get you closer

[Greg @ May 7th 2008 12:33PM]

uhm...neat? I do like the idea of a monster PC of course. Not sure they'll get anything useful for their money..but hey. Neat computer.

[radia @ May 7th 2008 12:33PM]

Wings, butterfly, ect.
1km? Try a differential volume element. Then you can accurately model the environment!
Lorenz must be rolling in his grave.

[steve @ May 7th 2008 12:34PM]

"That, they say, would be enough for it to create climate models at 1-kilometer scale"

How does this make sense. Is this in a day, a week, a year? You could model climate change on an GLOBAL scale with a 486 if you had enough time. (I think lol).

[radia @ May 7th 2008 12:36PM]

No, you can never model the environment on any scale for periods longer than about a week no matter how powerful or perfect your computer is, because it is a chaotic system.

[Johan S @ May 7th 2008 12:43PM]

Not necessarily, If you were to use a 486, it may take longer than the climate change itself.

That is lets say you want to simulate what the weather looks like in 100 years, witha 486 it may take 200 years to simulate 11 million years may take 2 million etc.

So I reckon to be declared as having the capability of computing climate change, you have to be able to compute it faster than the climate change itself.

It's possible that if you are computing the climate of 1 million years from now u dont need to simulate the intervening time .. but i'm ignoring that.

And yeah I do see your point .. it's valid etc. I am just being pedantic here.

[Joe @ May 7th 2008 2:48PM]

The scale refers to how detailed the models will be... or in n00b speak, how pixelized the output is.

[david @ May 7th 2008 12:38PM]

At this point, I thought the climate change thing was all but decided. How about we use all of the processing to cure some disease, or come up with materials or engineering designs that will allow us to do something about greenhouse emissions. Check this site: www.rmi.org

[digitallysick @ May 7th 2008 12:42PM]

Just enough power to run vista!

[radia @ May 7th 2008 12:44PM]

Climate change predictions based on computer models are pseudoscience. I'm not agreeing with the global warming denial crowd, but I think we're a long way from physically proving (a real proof, not some schoolhouse balloon experiment) that man-made greenhouse emissions are responsible. Please read up on Lorenz's work, and see why no matter how good your computer is, you're mathematically prevented from building the model. http://en.wikipedia.org/wiki/Edward_Lorenz

[Mike @ May 29th 2008 6:52AM]

20M CPU's will heat up the climate - easy to predict...

[TwinCityMike @ May 7th 2008 12:48PM]

Now researchers will be able to tweak their algorithms to get the predefined result they want even faster!

Global climate change has been happening for billions of years!

[Jandalf @ May 7th 2008 12:59PM]

So have UV radiation and tidal patterns and earthquakes. That doesn't mean we shouldn't study them to see how they do and will effect us - especially given how incredibly fast the human race has taken over this planet.

[phanbouy @ May 7th 2008 1:01PM]

idiots like you have been posting false equivalents on the internet for years as well, what's your point?

[Jandalf @ May 7th 2008 1:30PM]

If I'm the idiot you're referring to, phanbouy, my point was:

Climate change: environmental force that affects everyone.
UV radiation, earthquakes, tidal patterns: environmental forces that affect everyone.

So I'm in support of research to understand and prevent anthropogenic climate change on a large scale. Got it? Or am I still making false equivalents?

[phanbouy @ May 7th 2008 1:37PM]

not you Jandalf, Mike.

[Jandalf @ May 7th 2008 1:44PM]

Word.

[TwinCityMike @ May 7th 2008 2:05PM]

I am saying that many of these models are created and reverse engineered to give the result that is desired. and if you need the model to produce a certain result to get more funding, guess what will happen...

[phanbouy @ May 7th 2008 2:10PM]

that's right because why be an honest researcher when you can be a mad evil scientist who reverse engineers shit so that Al Gore can become world emperor and force everyone to drive Priuses.

[I LOVE THE CAPS LOCK KEY @ May 7th 2008 12:55PM]

20 million computers, that's a lot of Coal Power wasted.

[Low Ranked @ May 7th 2008 1:07PM]

20 million microprocessors, not computers.

[Xtort @ May 7th 2008 1:22PM]

COAL TASTES YUMMY!

[BigD145 @ May 7th 2008 3:09PM]

Berkley isn't part of the East Coast. They don't run on much coal.

[Urnamma @ May 7th 2008 1:08PM]

I wonder if Ray Kurzweil ever gets tired of being right.

[fuzzy @ May 7th 2008 1:00PM]

And since it was funded by Exxon Mobil, they will determine that climate change has nothing to do with man and that producing more oil and more greenhouse gases is in fact good for the environment.

[Andy TGD @ May 7th 2008 1:04PM]

Yeah fuzzy, cause like everyone knows that if you find a hole in the Man-made Global Warming theory you simply HAVE to be working with the oil companies. It couldn't possibly have anything to do with the fact that the world has be heating and cooling naturally for millions of years - never!

[phanbouy @ May 7th 2008 1:09PM]

wow.. you luddites are really in denial. go back and take a high school chemistry course before you post your middle school logic on why massively elevated levels of CO2 have no link whatsoever with climate change. freakin a

[willy the impeached @ May 7th 2008 1:19PM]

Close, but since it was funded by the feds it will continue to perpetuate imaginary warming senarios even as the earth cools.

I know it hurts but try reading something other than what the government tells you to.

[phanbouy @ May 7th 2008 1:29PM]

willy, you're a complete tool. you make the tinfoil hat crowd look nuanced

[TwinCityMike @ May 7th 2008 2:14PM]

willy, i think i am going to watch "inconvenient truth" tonight (again). I will then take it as 100% fact. I will then be an expert on this matter. This should help perpetuate the fear. If you want you can come over and watch it with me. but you have to ride your bike, i don't want your carbon footprint to get bigger because of me!

There seems to be a lot of experts on climatology here!

[phanbouy @ May 7th 2008 2:17PM]

twincityignoramous,

you are a really bad propagandist. before you malign the knowledge of others with your stupid foxnews talking point drivel, be aware that some of us have studied this issue in something called "universities" years before Al Gore popularized it. get over it.

[phanbouy @ May 7th 2008 2:30PM]

and by the way, these mystical places called universities have a process, check this out!, called "peer review". yes, "peer review" is a process whereby lots of scientists from different places and organizations give something what's called a venue for publishing. it tends to get more credibility than exxon-funded think tanks. can't imagine why.

[Andy TGD @ May 7th 2008 3:16PM]

phanbouy, yes you know everything. You are the most intelligent specimen on the face of this earth. Throwing around nice big words like "freakin a". With educated folks like yourself floating around it is no wonder ecomania has swept across the world! Please do not attempt to insult my intelligence with such inane drivel.

[phanbouy @ May 7th 2008 3:29PM]

wow andy, i must be some wacky eco-freak know-it-all because i mentioned peer review! come back when you have an argument.

[Andy TGD @ May 7th 2008 3:48PM]

No phanbouy, I was simply addressing the response which you made to me, personally. Actually, you may be surprised to find that I agree with your opinions regarding peer-review. But that standing, it still remains that the man-made global warming theory is simply one theory among many. I don't buy that theory, since there are, in my opinion, too many inconsistencies. Perhaps with your great wisdom, you can enlighten me to an answer for one of many gaps in my understanding of the theory - why is it that the temperature was rising prior to 1950 (during a period of relatively minor CO2 production) and then it dropped post 1950 during a time when one would expect CO2 levels to rise?

[TwinCityMike @ May 7th 2008 3:52PM]

phan...you hinted that you have studied this matter at a university. I would love to read your research on this matter. i am sure you did extensive research from an unbiased POV. universities are also driven by funding...in this world of honesty, no one would ever blow something out of proportion for more money and a secure job.

Also, i would like to see papers from the 90's that were written and studied, by this same mystical place you call a "university", about the hole in the ozone layer....that hole must be at least 10 times bigger by now....

[phanbouy @ May 7th 2008 3:57PM]

How many other theories are there? I haven't claimed to be a climate scientist, but I've studied the actual oceanographic and climatological carbon cycles in courses in soil science, oceanography, general biology, ecology, and other courses enough that I'm pretty convinced that anthropogenic greenhouse gas emissions can, have, and will continue to play a pretty influential part in the _accelerating_ rate of change in a) extremes in weather, b) desertification, c) sea level rise, etc. As far as I'm concerned the question of what _proportion_ of these changes is caused by greenhouse gas emissions (CO2, NOx, CH4, etc, etc) is small potatoes at this point. What is important is that we take steps to minimize new emissions and maximize sequestration of elevated atmospheric CO2. Why are humans so unconvinced of our own obvious ecological fragility?

[phanbouy @ May 7th 2008 4:01PM]

twincitymike, are you now also laying claim that the ozone hole is a lie? that is very very sad to live in such a paranoid little ignorant world. do some homework before making shit up, ok?

p.s. i claim you don't really exist and are an autopost bot funded by Exxon corp. you're guilty until proven innocent. k?

[phanbouy @ May 7th 2008 4:11PM]

btw, andy, it's pretty well documented that the lone "renegade" obligatory "skeptic" at any given university just happens to get his or her "research" funding directly from the Heritage Foundation. but I guess Mike would just say that any publicly funded research in the U.S. is designed to scare us about the environment because....uh.... the powers that be want us to be conservationists and not keep feeding record oil profits and buying useless shit. yep. that's it! good lord, at least get some consistency, people.

[o29 @ May 7th 2008 1:02PM]

If anyone even mentions Doom, I will punch you in the face.

[Esat Dedezade @ May 7th 2008 1:05PM]

You just did and that was almost as bad.

[o29 @ May 7th 2008 1:09PM]

I was 95% sure it would be mentioned at some point, and if it was going to be mentioned, I'd rather it be mentioned in a way that threatened those who keep the meme alive with physical violence. I'm not trying to keep it alive, and you'll never see me mention Doom again.

[w00t @ May 7th 2008 1:18PM]

But does it..... ugh.

Luckily I can't even bring myself to say it :)