The Seven Stones

What would you do with a petaflop?

The Blue Brain Project of the Ecole Polytechnique Fédérale de Lausanne (EPFL) currently uses an IBM Blue Gene/L supercomputer, reaching peaks of 22.4 teraflops (flops=Floating Point Operations Per Second), to simulate a model of a mammalian neocortical column composed of 10’000 neurons (Markram, 2006). IBM recently announced the release of the new Blue Gene/P, which, in its largest configuration, would be more than 100 times more powerful than the EPFL Blue Gene/L, reaching peaks of 3000 teraflops (3 petaflops, 3×1015 flops, 100’000 times more powerful than a home PC).

Apparently, this leap in computing power represents a challenge even for software developers, in particular due to the increase in parallelism (The petaflop challenge, Nature News). Would this open up fundamentally new avenues in systems biology? What are the applications in systems biology that would benefit the most from such supercomputing power? What would you do if you had a petaflop computer…?


Comments are closed.