AI-Driven Numerical Solver Research Scientist

NVIDIA; Washington (District of Columbia); On-site; USD 192,000 - 305,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA seeks an applied researcher to merge machine learning with numerical algorithms, accelerating solvers on modern GPUs. You will pursue AI-native methods such as AI-guided multigrid, differentiable solvers, and solver-in-the-loop learning to improve convergence, stability, and efficiency in large-scale simulations.

You’ll collaborate across NVIDIA Research, CUDA-X, and engineering teams, shaping the AI-native numerical methods agenda with evidence of impactful ML-to-solver connections.

Qualifications

• PhD or equivalent in a relevant field (computer science, ML, scientific computing, applied math, computational engineering, physics)
• 5+ years of relevant research or work experience
• Strong background in ML and scientific computing with demonstrated links to numerical algorithms
• Experience with PyTorch, JAX, or similar frameworks, plus Python and GPU computing
• Working knowledge of PDEs, sparse linear algebra, iterative solvers, preconditioning, FE/FV methods, optimization, or differentiable programming
• Research record in scientific ML, numerical linear algebra, or differentiable simulation integrated with solvers
• Strong communication skills for collaboration across AI research, numerical methods, product, and production teams

Responsibilities

• Research AI-assisted numerical methods to improve convergence, stability, accuracy, robustness, and wall-clock performance for large-scale simulations.
• Invent learned coarse spaces, learned preconditioners, AI-guided multigrid, solver-control policies, and hybrid numerical/ML algorithms.
• Build solver-in-the-loop pipelines using residual histories, discretized operators, meshes, geometry, and physics constraints.
• Define evaluation methods measuring convergence, failure rate, conservation, memory footprint, and end-to-end speedup.
• Collaborate across NVIDIA Research, CUDA-X, Warp, CAE, EDA, and related teams to advance AI-native numerical methods.
• Help define NVIDIA's applied research agenda for solver intelligence and AI-enabled numerical solvers

Skills

Machine learning

Numerical methods

PDEs

PyTorch/JAX

Python

GPU computing

Education

PhD or equivalent

Job description

NVIDIA seeks an applied researcher to merge machine learning with numerical algorithms, accelerating solvers on modern GPUs. You will pursue AI-native methods such as AI-guided multigrid, differentiable solvers, and solver-in-the-loop learning to improve convergence, stability, and efficiency in large-scale simulations.

You’ll collaborate across NVIDIA Research, CUDA-X, and engineering teams, shaping the AI-native numerical methods agenda with evidence of impactful ML-to-solver connections.