SciML Computational Scientist - Fusion Plasma Control

Userc Dehradun; Vaud; Hybrid; CHF 100,000 - 160,000; Full time

Benefits offered by this job

Immigration assistance

Equity compensation

Job summary

Fusionality in Lausanne is building the measurement, simulation, and control stack for fusion energy reactors. We seek a Computational Scientist specialized in Scientific Machine Learning to design differentiable solvers that run end-to-end on real plasma data and in control loops.

The role targets PhD-level or senior researchers with a strong background in SciML, ODE/PDE numerics, and differentiable frameworks, with production-grade performance on GPUs/TPUs.

Qualifications

• Substantial code in a differentiable framework (JAX, Julia, or PyTorch).
• Strong background in numerical methods for ODEs and PDEs including stiff and implicit solvers.
• Experience at the frontier of SciML, including Neural ODEs and hybrid models.
• Evidence of independent work: PhD or thesis, or code used by others.
• Nice-to-have: experience with Diffrax, Equinox, Optax, or Lineax.
• Familiarity with plasma/multiphysics simulation is a plus.

Responsibilities

• Design, implement and validate differentiable solvers in JAX for plasma physics problems such as free-boundary equilibrium, transport and state estimation.
• Make them robust and fast enough for production use via advanced numerical techniques for differential equation time stepping and GPU/TPU optimization.
• Use automatic differentiation for model calibration against experimental data, parameter estimation and controller design.
• Code ownership: quality, testing, verification against reference solutions, packaging and documentation.

Skills

SciML

Differentiable solvers

ODE/PDE numerics

JAX

PyTorch

GPU/TPU optimization

Automatic differentiation

Neural ODEs

Multiphyysics simulation

Education

PhD or equivalent research experience

Tools

Diffrax

Equinox

Optax

Lineax

Job description

Fusionality in Lausanne is building the measurement, simulation, and control stack for fusion energy reactors. We seek a Computational Scientist specialized in Scientific Machine Learning to design differentiable solvers that run end-to-end on real plasma data and in control loops.

The role targets PhD-level or senior researchers with a strong background in SciML, ODE/PDE numerics, and differentiable frameworks, with production-grade performance on GPUs/TPUs.