Mahesh Natarajan

Soonpil Kang, Ann S. Almgren, Mahesh Natarajan et al.

This paper describes an embedded boundary (EB) approach for simulating three-dimensional fluid flow on a staggered mesh where the velocity components are defined on cell faces and the thermodynamic state is defined on cell centers. Most EB approaches assume that all components of the solution, including the velocity, are co-located. To compute solution quantities on faces as well as cell centers, we construct and store multiple instances of the geometric information, one for the quantities stored at cell centers and one for each velocity component. In addition, we extend the weighted state redistribution (WSRD) scheme to staggered meshes to address the small-cell instability issue. This new approach is implemented in the Energy Research and Forecasting (ERF) model that provides performance portability and adaptive mesh refinement. We validate the new EB method by comparing EB simulations to those computed using terrain-following coordinates.

Mahesh Natarajan, Xiaoye Li, Weiqun Zhang

We present AstraAI, a command-line interface (CLI) coding framework for high-performance computing (HPC) software development. AstraAI operates directly within a Linux terminal and integrates large language models (LLMs) with Retrieval-Augmented Generation (RAG) and Abstract Syntax Tree (AST)-based structural analysis to enable context-aware code generation for complex scientific codebases. The central idea is to construct a high-fidelity prompt that is passed to the LLM for inference. This prompt augments the user request with relevant code snippets retrieved from the underlying framework codebase via RAG and structural context extracted from AST analysis, providing the model with precise information about relevant functions, data structures, and overall code organization. The framework is designed to perform scoped modifications to source code while preserving structural consistency with the surrounding code. AstraAI supports both locally hosted models from Hugging Face and API-based frontier models accessible via the American Science Cloud, enabling flexible deployment across HPC environments. The system generates code that aligns with existing project structures and programming patterns. We demonstrate AstraAI on representative HPC code generation tasks within AMReX, a DOE-supported HPC software infrastructure for exascale applications.