Computational Scientist, Differentiable Physics

Sourced from zerogtalent.com

Periodic Labs is seeking a Computational Scientist to build differentiable, accelerator-ready simulations for continuum-physics problems. You will combine governing equations, numerical methods, and deep learning to create scalable surrogate models and optimize physical systems. The role involves developing multi-scale solvers, validating against benchmarks, and leveraging frameworks like JAX or PyTorch.

• Build and extend differentiable solvers for continuum simulation and multi-physics problems
• Implement numerical methods from equations and diagnose convergence and stability
• Combine simulation with deep learning for surrogate modeling and optimization
• Validate models against experiments and high-fidelity simulations
• Create datasets and evaluations to guide LLM development for scientific automation

• PhD or equivalent research experience in applied mathematics, computational science, physics, or engineering
• Code-level experience building or modifying PDE solvers and differentiable simulations
• Deep expertise in at least one continuum domain (e.g., fluid dynamics)
• Experience training and evaluating deep-learning models for physical systems