Vinci4d
Software Engineer – Nonlinear Solid Mechanics & High-Performance Computing
2w ago
190000 –230000 USD / yearUSAMiddleRemotec++high-performance computingnumerical methodsci/cdnonlinear mechanics
Develop and optimize nonlinear solvers for solid mechanics in high-performance computing environments.
About the company
- At Vinci4d, we are building the next generation of simulation software for thermal, fluid flow, and structural mechanics applications — the kind of tools that change how engineers design products, from the first mesh to the final answer. We are a small, technically deep team that moves fast, ships real software, and takes on hard problems that matter. If you want your work to be foundational to a platform used by engineers worldwide, this is the place.
Responsibilities
- We are looking for a software engineer who lives at the intersection of computational solid mechanics, numerical methods, and high-performance computing. You will design, implement, and tune solvers for geometric and material nonlinearity in solid mechanics — think large-deformation, contact, and history-dependent material response — that run at scale on modern hardware. You will write production-quality code, contribute to our CI/CD infrastructure, and collaborate closely with a multi-disciplinary team of physicists, engineers, and software developers.
- This is not a "maintain the existing stack" role. You will be building things that don't exist yet, solving problems that require both rigorous mathematical thinking and solid engineering instincts.
Nice to have
- Experience with warpage and residual-stress problems in semiconductor manufacturing (e.g., packaging, die/substrate stacks, thermomechanical deformation)
- Familiarity with matrix-free methods for nonlinear and linear operator application
- Experience with geometric multigrid approaches as solvers or preconditioners
- Background in adaptive mesh refinement (AMR)
- Familiarity with embedded geometry or immersed boundary methods for solid mechanics
- Experience applying machine learning to solid mechanics problems (surrogates, constitutive modeling, solver acceleration)
- Experience with performance profiling tools (Nsight, VTune, Roofline analysis)
Other
- Develop and tune nonlinear solvers for solid mechanics, handling both geometric nonlinearity (large deformation, finite strain) and material nonlinearity (plasticity, viscoelasticity, temperature-dependent and history-dependent constitutive models)
- Build and optimize the underlying linear algebra: iterative linear solvers and preconditioners for the large sparse systems arising at each Newton iteration
- Port and optimize these solvers for GPU execution using CUDA, HIP, or equivalent frameworks, with a focus on memory bandwidth, occupancy, and scalability
- Implement FEM discretizations for structural and thermomechanical field solves, with attention to robustness and convergence under stiff, ill-conditioned, and near-singular conditions
- Contribute to a robust software engineering foundation: version control discipline, automated testing, CI/CD pipelines, and code review practices
- Collaborate with domain experts to translate physical models and mathematical formulations into correct, efficient implementations
- Profile and benchmark solver performance; identify and eliminate bottlenecks
- Hands-on experience developing solvers for geometric and material nonlinearity in solid mechanics — large-deformation kinematics, nonlinear constitutive models, and the Newton-type schemes that drive them to convergence
- Strong foundation in the finite element method (FEM) for solid and structural mechanics
- Deep familiarity with iterative linear solvers (e.g., Krylov methods) and preconditioning techniques for large, sparse systems, with hands-on experience implementing these inside a nonlinear solver
- Proven GPU programming experience (CUDA, HIP, SYCL, or similar) with a track record of getting real performance out of hardware
- Proficiency in C++ and/or Python; comfort working in performance-critical codebases
- Strong software engineering practices: Git workflows, code review, automated testing (unit, integration, regression), and CI/CD pipelines
- 3–6 years of industry or research experience in a relevant field (computational mechanics, scientific computing, computational physics, numerical simulation, or HPC)
- A portfolio of work — open source contributions, published code, or shipped products — that demonstrates the above
- A genuine collaborator: you learn from teammates as readily as you help them
- Able to communicate technical depth clearly to people from different disciplines — physicists, mechanical engineers, product managers
- Comfortable with ambiguity and excited by the challenges that come with building something new
- Self-directed and ownership-oriented: you drive your work to completion without needing to be managed closely
- Work on genuinely hard technical problems with real engineering impact
- Join a small team where your contributions are visible and your voice is heard
- Competitive compensation with equity participation
- Flexible work environment
- The satisfaction of building something from the ground up — and the opportunity to help define what it becomes