Computational Scientist, Differentiable Physics

Periodic Labs seeks a Computational Scientist to develop differentiable physics simulators for continuum problems, combining numerical methods, deep learning, and modern accelerators. The role involves building scalable solvers, implementing physics-informed models, and creating evaluation frameworks to advance AI-driven scientific simulation.

Responsibilities

• Build and extend differentiable solvers for continuum simulation including multi-scale and multi-physics problems
• Implement numerical methods from equations and literature, diagnosing convergence and stability issues
• Combine simulation with deep learning for surrogate modeling, inverse problems, and optimization
• Validate models against experiments, benchmarks, and high-fidelity simulations
• Create datasets and evaluation frameworks to guide LLM development for task automation

Requirements

• PhD or equivalent research experience in applied mathematics, computational science, physics, engineering, or related field
• Code-level experience building or modifying PDE solvers, numerical methods, or differentiable simulations
• Deep expertise in continuum physics domain with ability to learn new physics rapidly
• Experience training and evaluating deep-learning models for physical systems
• Strong Python proficiency with JAX, PyTorch, Julia, or C++
• Applied simulation experience on realistic scientific or engineering problems
• Startup mentality with ownership, judgment under uncertainty, and ability to build independently

Compensation: $250,000-$350,000 base + equity