GPU Computing Specialist - LAVA CFD Team

Analytical Mechanics Associates•Mountain View, CA

35d

About The Position

Analytical Mechanics Associates (AMA) is seeking a skilled and experienced GPU Computing specialist to support the Launch Ascent and Vehicle Aerodynamics (LAVA) team within the Computational Aerosciences Branch (TNA) at NASA Ames Research Center (ARC) under the Aircraft Systems and Speceflight Engineering Support Servies (ASSESS) contract. The successful candidate will play a key role in advancing GPU acceleration, performance optimization, and feature development of the LAVA CFD solver suite — a high-performance computational fluid dynamics (CFD) solver framework used in mission-critical aerospace simulations.

Requirements

Advanced degree in Computational Sciences or related field
Experience with CUDA programming and GPU architecture optimization
Knowledge of detailed CUDA optimization techniques (occupancy, registers, memory layout, etc.), and profilers (ncu and nsys)
Strong C++ programming skills with experience in performance-critical applications
Working knowledge of MPI and OpenMP parallel programming models
Demonstrated ability to optimize scientific computing applications
Candidates must have current U.S. work authorization. While preference is given to U.S. citizens and lawful permanent residents, we also welcome applications from individuals holding valid U.S. work visas (such as OPT, CPT, or other permissible categories).

Nice To Haves

Background in computational fluid dynamics or related scientific computing fields
Knowledge of GPU computing on NVIDIA HPC architectures (eg A100, GH200)
Experience with octree data structures and Cartesian AMR algorithms
Experience working in diverse computing environments from workstations to HPC systems
Knowledge and experience with ray tracing libraries (OptiX, OWL, Embree, etc.) for immersed boundary geometry queries
Knowledge of FEM structural dynamics for fluid-structure interaction simulations
Experience with performance portability layers
Experience with non-NVIDIA GPU accelerators (AMD, Intel, etc.)
Familiarity with cross-platform GPU development
Familiarity with low-dissipation numerics for computational fluid dynamics
Experience with multi-phase or multi-physics simulations
Background in aerospace applications, particularly: Launch environment modeling, Parachute/entry system dynamics, Aerodynamic simulations of aircraft

Responsibilities

Contribute to GPU optimization efforts for the LAVA Cartesian solver, focusing on MPI/C++/CUDA implementation
Support enhancement of Wall-Modeled Large Eddy Simulation (WMLES) capabilities
Implement and optimize new features for the adaptive mesh refinement framework
Assist with hybrid parallelization strategies using MPI, CUDA, and OpenMP
Help identify and resolve performance bottlenecks in computational workflows
Create comprehensive technical documentation for code modifications and new features
Assist users with software issues as we release the LAVA code to the broader community
Collaborate with multidisciplinary scientific teams to address complex aerospace CFD problems
Support simulation deployment across various computing environments, from workstations to supercomputers