Senior Applied Research Scientist – GPU Native Numerical Algorithms
NVIDIAJob Title
Senior Applied Research Scientist – GPU Native Numerical Algorithms
Role Summary
Join NVIDIA's computational engineering applied research team to design GPU-native numerical methods for engineering simulation. The role focuses on inventing and evaluating algorithms and research prototypes tuned for modern GPU architectures and multi-GPU systems.
The work spans numerical analysis, accelerated computing, and production-minded software engineering; outcomes move promising ideas toward NVIDIA software used by researchers, engineers, and partners.
Experience Level
Senior - typically expects 5+ years of relevant work or research experience.
Responsibilities
Contribute to research and prototype development of GPU-native numerical algorithms and solver technologies.
- Invent and reformulate numerical algorithms co-designed for NVIDIA GPU architectures (implicit and explicit simulation).
- Develop and evaluate linear and nonlinear solver approaches: Newton–Krylov, multigrid/AMG, domain decomposition, matrix-free methods, mixed precision, sparse iterative/direct solvers, and preconditioners.
- Investigate alternatives to CPU-oriented methods for GPUs, including synchronization-avoiding Krylov, GPU-native multigrid, and mixed-precision/tensor-core approaches.
- Benchmark and validate algorithms on representative workloads across mechanics, thermal-fluid, electromagnetics, semiconductor and EDA domains.
- Collaborate with CUDA-X, Warp, solver engineering, NVIDIA Research, universities, and industry partners to transition prototypes toward production software.
- Help shape the applied research roadmap for GPU-native and AI-native computational engineering methods.
Requirements
Must-have technical skills and experience for immediate impact; followed by concise nice-to-have items.
- 5+ years of relevant work or research experience in computational science or engineering domains.
- Research or engineering experience with PDE discretization and numerical methods: finite element, finite volume, discontinuous Galerkin, nonlinear solvers, sparse linear algebra, and preconditioning.
- Proficient programming in C++ and Python and experience developing or optimizing CUDA/GPU code.
- Experience using profiling, benchmarking, and numerical validation to improve algorithm performance on GPU or multi-GPU systems.
- Clear technical communication and demonstrated ability to collaborate across research, engineering, product, and partner teams.
Nice-to-have:
- Experience with implicit structural dynamics, nonlinear mechanics, contact, CFD, electromagnetics, multiphysics, or semiconductor/EDA simulations.
- Familiarity with PETSc, Trilinos, MFEM, libCEED, OpenFOAM, NVIDIA Warp, CUDA-X libraries (cuSPARSE, cuSOLVER), or similar frameworks.
- Experience with industrial CAE/EDA tools and CAD-connected workflows (Ansys, Abaqus, LS-DYNA, Siemens, Dassault, Altair, Cadence, Synopsys, COMSOL, MathWorks).
- Experience with distributed solvers using MPI/NCCL, asynchronous methods, or GPU-cluster performance analysis.
- Publications, patents, open-source contributions, or deployed computational science software.
Education Requirements
PhD or equivalent experience in computational mechanics, applied mathematics, scientific computing, computer science, aerospace, mechanical, civil engineering, or a related technical field (the posting explicitly allows equivalent practical experience).
About the Company
Company: NVIDIA
Headquarters: Santa Clara, California, USA
NVIDIA is a global leader in accelerated computing, renowned for its innovative solutions in AI and digital twins that transform diverse industries. The company specializes in networking technologies, providing end-to-end InfiniBand and Ethernet solutions for servers and storage that optimize performance and scalability. NVIDIA serves sectors such as high-performance computing, enterprise data centers, and cloud computing, constantly reinventing its products and services to stay ahead in the market.
