The National Ignition Facility (NIF) [1] is a high energy density experimental facility run for the National Nuclear Security Administration (NNSA) by Lawrence Livermore National Laboratory. NIF houses the world’s most powerful laser. The intensity of the NIF lasers beams is high enough (over 1014W/cm2) that interactions between the laser beams and fluctuations in the density of the ions and electrons can scatter significant amounts of laser light.

pF3D [2], [3] is a laser-plasma interaction code used to assess proposed experimental designs for expected levels of scattering and to help understand measurements of scattered light in NIF experiments. NIF experiments have shown that laser-plasma interactions transfer significant amounts of energy between beams and increase the amount of backscattered light relative to what would occur without energy transfer.

Single beam pF3D simulations have been run for several years and can be used in regions where the beams are significantly sepa- rated. We have recently begun running multi-beam simulations to understand the scattering in the regions where the beams overlap. These multi-beam simulations may require over 100 billion zones and typically run for 1-4 weeks. The Cielo computer at Los Alamos National Laboratory is the only NNSA computer with enough memory and compute performance to run the multi-beam simulations we need to understand key experiments currently being carried out with the NIF laser.

This talk reports on the approaches we took on Cielo to achieve the high I/O and message passing rates necessary to run full system pF3D simulations efficiently. We compare and contrast the approaches that work well on a Cray XE 6 with a Gemini interconnect and a Panasas parallel file system to the approaches we took in previous simulations on BlueGene/P systems with Lustre and GPFS parallel file systems.