Wednesday, November 19, 2008

Supercomputer Hyperbole

Wired has an article about the latest iteration of supercomputers breaking the petaflop barrier. That means they can carry out just over a quadrillion (1,000,000,000,000,000) floating-point calculations per second. That's a lot. And it's a significant milestone. But this seems like a bit much:

"The scientific method has changed for the first time since Galileo invented the telescope (in 1509)," said computer scientist Mark Seager of Lawrence Livermore National Laboratory.

Look, this is a quantitative, not a qualitative, change. Some are making the claim that the ability to model and simulate at greater and greater levels of detail will allow for qualitatively different ways of doing science. But I don't think so.

Breaking the petaflop barrier is more like building a bigger and better telescope that allows us to see farther and clearer, and to see things we've never seen before, but it's not like the invention of the telescope in the first place.

No comments: