Publications

Abstract not provided.

Activities, accomplishments, next steps and outreach are reported, primarily related to the Kokkos project.

Sections include: performance to plan, components, exceeds, and lessons learned. Updated areas include: SPARC, UMR, EMPIRE, Panzer.

Abstract not provided.

Abstract not provided.

The SPARC team has completed much of the workflow development needed for the coupled and uncoupled analyses needed to support our flight test validation work. The SPARC team has continued to make progress on development activities to support unsteady, full reentry vehicle aero analysis, which have focused on turbulence modeling, uniform mesh refinement, in-situ visualization (with the Paraview/Catalyst team), and performance analysis. The SPARC team is making solid progress towards our Q4 goal of demonstrating an improved Wall Modeled Large Eddy Simulation (WMLES) capability for standard geometries using high-order finite difference, discontinuous Galerkin, and low-dissipation finite volume methods. The SPARC team has met the minimum completion criteria for parts of the L1, including documenting performance and scaling of SPARC on Trinity, Sierra and Astra, and performing runs for the flight test validation case.

Abstract not provided.

The SPARC and SPARC V&V teams successfully presented their work at the Dec 11-12 L1 milestone mid-year review. The teams received overwhelmingly positive feedback at the review.

Supporting the latest hardware and compiler versions is important to leverage improvements in the software environment and new HPC platforms. We will provide certified support for the latest releases of vendor compilers from Intel, AMD, IBM, NVIDIA, ARM and Cray as well as of open source compilers GCC and Clang.

Provide high quality (production) Kokkos support and consultation for ASC applications and libraries.

Engage the C++ standards committee to further the adoption of successful Kokkos concepts into the C++ standard, and provide feedback on proposed concurrency mechanisms such as the executors proposal.

Harden and optimize the ROCm based AMD GPU backend, develop a prototype backend for the Intel ECP Path Forward architecture, and improve the existing prototype Remote Memory Space capabilities.

Abstract not provided.

Accomplishments and progress are reported for quarter 3, FY2019.

The four key emission models necessary for B-dot cavity physics and DSMC (Direct Simulation Monte Carlo) collision models for Argon background gas have been implemented and tested. A series of B-dot simulations have been performed with vacuum, gas fill, and gas fill with space-charge limited emission models. Results are being validated against NIF experimental data. These were presented at the mid-term panel review. The ATS-1 L2 milestone is dependent on this milestone and is similarly on track. The mid-term review occurred at SNL on 5/15/2019. Target problems at the requisite scale and of interest to this Milestone have been defined and will be run during Q4.