This paper examines whether lossy compression can be used effectively in physics simulations as a possible strategy to combat the expected data-movement bottleneck in future high performance computing architectures. This paper shows that, in a number of cases, compression levels of 3-5X can be applied without causing a significant change in the physical quantities that are of most interest for each simulation. Rather than applying classical error metrics from signal processing, we utilize

more »

... s-based metrics appropriate for each code to evaluate the impact of compression. We evaluate simulations run with three different codes: a Lagrangian shock-hydrodynamics code, an Eulerian higher-order hydrodynamics turbulence modeling code, and an Eulerian coupled laser-plasma interaction code. We apply compression to relevant quantities after each time-step to approximate the effects of tightly coupled compression and also study the compression rates to estimate memory and disk-bandwidth reduction. We find that the error characteristics of compression algorithms must be carefully considered in the context of the underlying physics being modeled.