Even with "such a rich, important simulation" behind them, researchers still have more uses for this three-dimensional adaptive mesh refinement cosmology code.

First of all, they haven't fully simulated that first star yet - only the clouds that will give birth to them.

"The simulation has not actually been carried out to the formation of a star, but the code is capable of it. And we're damn close. Close enough to say 'that thing is going to become a star,'" Norman says. "We've made a protostellar cloud core that's collapsing like a freight train." The code is capable of recreating this collapse, and it's only a matter of time before the simulations catch up with the extrapolation.


Perhaps more exciting, however, is the variety of cosmology problems that the code can be used to solve. With only minor alterations that "fit very nicely into the code," according to Bryan, it can also simulate groups of stars as gravity pools them into galaxies. Currently, Bryan is even using the code on 128 processors of NCSA's Origin2000 to study groups of galaxies as they clump into galactic clusters.

The scope of problems that can be studied make the code "a real giant killer," Norman says. "In many ways, we've got a key to the universe."

1 2 3 4 5