Accelerated Physics Simulation Engineer (GPU-Driven AI)

Industrious Ventures; Los Angeles, Northern (CA, KY); Hybrid; USD 165,000 - 250,000; Full time

Job summary

Voyager Technologies seeks an Accelerated Physics Simulation Engineer to build fast, high-fidelity physics simulations for AI-driven hardware design exploration. You’ll develop differentiable models and surrogate solvers, validating them against high-fidelity tools and real data, spanning PDEs, GPUs, and ML surrogates.

The role sits in the ACE team within Advanced Technology Development, designing an AI-native platform that compresses hardware cycles from years to days.

Qualifications

• PhD in Computational Physics, Mechanical or Aerospace Engineering, Applied Mathematics, Computer Science (numerical methods) or related field; or Master’s +3 years.
• 0-3 years post-PhD industry/startup/postdoc or 3-6 years in computational science/engineering.
• Hands-on experience implementing numerical methods for PDEs (FEM, FVM, FDM, particle/mesh-free) in research or production.
• Experience with at least one ML framework (JAX, PyTorch, TensorFlow) and GPU tech (CUDA, PhysicsNEMO).
• Experience speeding up simulations or building surrogate models with quantitative results.
• Experience using LLMs to accelerate boilerplate coding and experiments.

Responsibilities

• Design fast physics solvers (CFD, thermal, structural, plasma) for agentic optimization loops.
• Develop surrogate models (PINNs, neural operators, graph nets) with large cost reductions.
• Integrate solvers and surrogates into the ACE platform for AI agents to call as tools.
• Collaborate with ACE leads to identify regimes and validate models.
• Create datasets by coupling commercial/open-source solvers with parameter sweeps.
• Profile and optimize GPU kernels and pipelines for speedups with maintained accuracy.
• Develop test harnesses, benchmarks & diagnostics for models over time.
• Use LLMs to accelerate boilerplate coding and scripting.
• Leverage advanced LLMs and tooling for complex mathematics and simulations.

Skills

PDE methods

GPU acceleration

LLM-assisted coding

Education

PhD in Computational Physics or related

Master's + 3 years experience

Tools

CUDA

JAX

PyTorch

TensorFlow

OpenFOAM

Job description

Voyager Technologies seeks an Accelerated Physics Simulation Engineer to build fast, high-fidelity physics simulations for AI-driven hardware design exploration. You’ll develop differentiable models and surrogate solvers, validating them against high-fidelity tools and real data, spanning PDEs, GPUs, and ML surrogates.

The role sits in the ACE team within Advanced Technology Development, designing an AI-native platform that compresses hardware cycles from years to days.