Differentiable Simulation & AI Research Intern

NVIDIA Corporation; Santa Clara (CA); In loco; USD 52.000 - 129.000; Tempo pieno

Vantaggi offerti da questo lavoro

Intern benefits

Descrizione del lavoro

NVIDIA Corporation is seeking a PhD candidate for an internship focused on scalable differentiable algorithms. Work spans mathematics, code, hardware, and engineering applications, with emphasis on adjoint and automatic differentiation methods in GPU-accelerated simulations.

You will prototype GPU-native solvers, integrate AI-based methods with simulations, and evaluate accuracy, convergence, and scalability across CFD, mechanics, and multiphysics workloads.

Competenze

• Pursuing a PhD in applied mathematics, scientific computing, computational engineering, CS, or related field.
• Strong foundations in PDEs, numerical linear algebra, iterative solvers, optimization, or scientific computing.
• Experience with adjoint methods, automatic differentiation, inverse problems, or data assimilation.
• Experience writing numerical software in Python and C++, with GPU computing experience.
• Ability to design experiments, compare numerical tradeoffs, and communicate results clearly.

Mansioni

• Prototype GPU-native solvers with adjoint and autodiff techniques for engineering simulations.
• Integrate AI-based methods with simulations for parameter calibration and data assimilation.
• Co-design algorithms for GPU parallelism, memory, mixed precision, and multi-GPU execution.
• Evaluate numerical accuracy, convergence, gradient quality, and scalability on CFD and multiphysics workloads.
• Move promising methods toward reusable capabilities in CUDA-X libraries.

Conoscenze

Python

C++

CUDA

GPU Computing

Numerical Methods

Formazione

PhD candidate in applied mathematics / scientific computing

Strumenti

CUDA-X libraries

PyTorch

JAX

OpenFOAM

Descrizione del lavoro

NVIDIA Corporation is seeking a PhD candidate for an internship focused on scalable differentiable algorithms. Work spans mathematics, code, hardware, and engineering applications, with emphasis on adjoint and automatic differentiation methods in GPU-accelerated simulations.

You will prototype GPU-native solvers, integrate AI-based methods with simulations, and evaluate accuracy, convergence, and scalability across CFD, mechanics, and multiphysics workloads.