Steve Chen Making China's Supercomputer Grid

nanotrends writes "Steve Chen was the principal designer of the Cray X-MP supercomputer. He recently created multi-teraflop blade based supercomputers for a Chinese company. He is now creating a supercomputer grid across China and he is working on a bio-supercomputer extension to human brains called THIRD-BRAIN. The THIRD-BRAIN project has significant 3 year and 5 year targets."

Comments

I certainly hope this plan is actually reasonable.

[neoshroom]

It isn't. From the article, the guy right now has nothing and plans to implement what the entire AI field has been trying to do for years and years. He doesn't mention any new or innovative techinques he is applying to the problem -- only a boatload of unfounded confidence.

One brain in your head, one in your pants

[macadamia_harold]

and he is working on a bio-supercomputer extension to human brains called THIRD-BRAIN. The THIRD-BRAIN project has significant 3 year and 5 year targets.

Why call it Third Brain? I'm not terribly familiar with Anatomy, but I thought humans only had one brain. (Well, unless you're a guy. Then you've got an extra one in your pants that does most of the thinking)

Haha. Not funny.

[Anonymous Coward]

s/China/US/
s/communist/capitalist/
s/Tibet/Iraq/

You cannot even talk about their dumb leaders after electing Bush.

Maybe someone comes up with moral enough to make speeches. This is not America, though.

Strange...

[Shaltenn]

Strange... Call me crazy, but this reeks of Cyber-brains from Ghost in the Shell. XP

Re:Difficult to take 3rd brain seriously

[KillerCow]

He is now creating a supercomputer grid across China and he is working on a bio-supercomputer extension to human brains called THIRD-BRAIN.

But when queried about the AI aspect, he says that they will just have to research it, even after the interviewer points out many other people have been working on this. I think he greatly underestimates the problem! Saying that all other AI researchers have not had an integral approach is a bit lacking as an answer to me.

It's pure hubris. Even if he does make some AI breakthrough (unlikely), he may find that his shiny new grid is not suited to its implementation. It may not clique well. He has a system (a grid) and is trying to apply it to a known problem (AI). It's a lofty and noble goal, but it isn't the best way to solve a problem. Usually, it's a good idea to have your algorithms and systems though out (or at least an idea of what they are) before starting implementation.

Fortunately for him though, he has a nice preexisting model that he can copy the structure of. Our brains are distributed, and are cliqued (from my understanding). So building a grid and trying to apply it to AI is not a terrible idea.

At least the Chinese will have a nice super-grid to do weapons research on in the mean time.

Re:I certainly hope this plan is actually reasonab

[RAMMS EIN]

``only a boatload of unfounded confidence.''

Well, isn't that what made America so great?

Re:A Great Leap Forward in computing?

[An Onerous Coward]

Might I add that this abdication has turned out to be wildly successful?

Not universally. In health care, our mostly private system is the most expensive in the world, yet we have about the same life expectancy as Cuba (with a higher infant mortality rate). In agriculture, we overproduce and overconsume things that are frankly not good for us. Our food is overprocessed, our diets too high in lard and sugar, and we eat far more than is healthy (in no small part thanks to the unceasing, mind-warping advertisements plastered on every visible surface in our country).

This is in no small part due to government agricultural subsidies, but the fact that these subsidies continue despite the fact that they actively undermine our health has to say something about the power our government grants business.

Meanwhile, the wages of the poorest 40% of Americans have hardly budged since the 1970s, while the richest 1% of Americans now own about half the wealth in this country. Is that the sort of economy anyone would plan, or anyone would ask for? Besides the aforementioned 1%, I mean.

In a way, we have the worst sort of mix of private and public: a government which allows business to write its rules. Capitalism is no more a panacea than communism. Global capitalism isn't all prosperity and health and flat-screen TVs; it also brings wage arbitrage, huge wealth inequalities, unsustainable overconsumption, sweatshop labor, and a host of other problems. In your rush to defend the good aspects of the free market, you shouldn't ignore its pitfalls and limitations.