Senior GPU-Native Numerical Methods Scientist

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

Benefits offered by this job

Equity and benefits

Job summary

NVIDIA is seeking an Applied Research Scientist to design GPU-native numerical methods for engineering simulation and to build research prototypes that scale across multi-GPU systems. You will explore solver algorithms, evaluate them on engineering workloads, and collaborate with CUDA-X, NVIDIA Research, and industry partners to move ideas into production software.

Applicants should hold a PhD or equivalent with 5+ years of experience in computational mechanics or scientific computing, with

Qualifications

• PhD or equivalent experience in computational mechanics, applied mathematics, scientific computing, or related field.
• 5+ years of relevant work/research experience.
• Experience with PDE discretization, finite element/volume, nonlinear solvers, sparse linear algebra, or high-performance computing.
• Experience writing numerical software in C++ and Python, with CUDA or GPU code.
• Experience with profiling, benchmarking, or performance analysis on GPU or multi-GPU systems.

Responsibilities

• Invent and reformulate numerical algorithms for modern NVIDIA GPU architectures.
• Develop linear and nonlinear solver approaches (Newton-Krylov, multigrid, domain decomposition).
• Investigate GPU-native reformulations for existing CPU methods.
• Evaluate algorithms on engineering workloads (mechanics, thermodynamics, electromagnetics).
• Collaborate with CUDA-X, NVIDIA Research, universities, and industry partners to move research to production software.

Skills

CUDA

C++

Python

GPU programming

Numerical methods

Education

PhD or equivalent in computational mechanics / scientific computing

Tools

CUDA

Job description

NVIDIA is seeking an Applied Research Scientist to design GPU-native numerical methods for engineering simulation and to build research prototypes that scale across multi-GPU systems. You will explore solver algorithms, evaluate them on engineering workloads, and collaborate with CUDA-X, NVIDIA Research, and industry partners to move ideas into production software.

Applicants should hold a PhD or equivalent with 5+ years of experience in computational mechanics or scientific computing, with