GPU-Native Numerical Methods Scientist

NVIDIA; Massachusetts; On-site; USD 192,000 - 357,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA is seeking an Applied Research Scientist to advance GPU-native numerical methods for engineering simulation across NVIDIA platforms. You will design, prototype, and compare solver algorithms, focusing on massive parallelism, mixed precision, and multi-GPU systems.

The role blends numerical analysis, high-performance computing, and production-minded software engineering, requiring collaboration with CUDA-X, Warp, and NVIDIA Research to move ideas into software used by researchers and

Qualifications

• PhD or equivalent experience in computational mechanics, applied mathematics, scientific computing, CS, or related field.
• 5+ years of relevant work/research experience.
• Experience with PDE discretization, finite element/finite volume, and nonlinear solvers.
• Experience writing numerical software in C++ and Python; CUDA or GPU code experience.

Responsibilities

• Invent and reformulate numerical algorithms co-designed for modern NVIDIA GPU architectures.
• Develop linear and nonlinear solvers including Newton-Krylov, multigrid, and preconditioning.
• Investigate GPU-native reformulations of CPU-oriented methods and mixed-precision algorithms.
• Evaluate algorithms on workloads across CAE domains and semiconductor simulation.
• Collaborate with NVIDIA teams and academic partners to move research toward production software.

Skills

PDE discretization

Finite element / Finite volume

Nonlinear solvers

Sparse linear algebra

High-performance computing

Communication of technical tradeoffs

Education

PhD or equivalent experience in computational mechanics / scientific computing

Tools

CUDA

Python

C++

Job description

NVIDIA is seeking an Applied Research Scientist to advance GPU-native numerical methods for engineering simulation across NVIDIA platforms. You will design, prototype, and compare solver algorithms, focusing on massive parallelism, mixed precision, and multi-GPU systems.

The role blends numerical analysis, high-performance computing, and production-minded software engineering, requiring collaboration with CUDA-X, Warp, and NVIDIA Research to move ideas into software used by researchers and