Simulating N-body systems on the grid using dedicated hardware

We present performance measurements of direct gravitational N-body simulation on the grid, with and without specialized (GRAPE-6) hardware. Our intercontinental virtual organization consists of three sites, one in Tokyo, one in Philadelphia and one in Amsterdam. We run simulations with up to 196608 particles for a variety of topologies. In many cases, high performance simulations over the entire planet are dominated by network bandwidth rather than latency. Using a global grid of GRAPEs our calculation time remains dominated by communication over the entire range of N, which was limited due to the use of three sites. Increasing the number of particles will result in a more efficient execution and for N >= 2 · 10 to the power of 6 we expect the computation time to overtake the communication time. We compare our results with a performance model and find that our results are in accordance with the predicted values.