ASCI White Detonates The First E-Bomb

totallygeek writes "Redefining the term vaporware, research scientists at Lost Alamos and Lawrence Livermore Labs detonated two computer simulations. ASCI White, the world's fastest supercomputer, ran the simulations of nuclear explosions. Scientists can now study nuclear weapon replacement components without violating the nuclear test ban, in effect since 1992. Each simulation used more than 6.6 million CPU hours, which would take home machines 1000 years to complete. The data for each experiment was equivalent to 35 times the information available in the Library of Congress. ASCI White currently operates at 12 teraflops, but by early next year, Los Alamos expects to operate at 30 teraflops. The seven month research project ended last Friday, and now the system is ready for use, after its sucessful testing."

Comments

Re:And this benefits mankind how?

[NanoGator]

"Why isn't this processor farm being used for, say, bacterial genomics?"

You'd rather have potentially defective nuclear weapons at our disposal than advancing bacteria research?

I can see it now: "The United States was forced to launch a nuclear assault on Iraq today. Unfortunately, the first bomb that went off was powerful enough to scare birds away. In live television address, Saddam Hussein's speech was surprisingly short. 'Ha ha!', he said as he pointed in the general direction of the United States."

There's a pretty big difference between modelling a nuclear explosion (or an explosion of any kind) and performing Seti like research. For one thing, Seti can divvy up the data to process into really small chunks, making it easy to distribute it across a number of machines.

This cannot be done with an explosion of any kind. The reason is that you have LOTS of particles interacting with each other. For each interaction, every single particle needs to be re-calculated. This is why you cannot divvy up the data and spread it across a lot of machines. This is why you need to use a computer like this to do the calculation.

I'm not sure what kind of math is done for genome related stuff, so I cannot really comment on that. I'm willing to bet, though, they could spread it across as many machines as they need. If they need a machine like the one being used in this article, they could probably lease run time off it.