PhD Intern, Computational Physics-Shape Optimization

About the Role

Interns will contribute to the SOS Group’s technical expertise by conducting original research in shape sensitivity, structural optimization, and shape parameterizations. They will derive sensitivity equations for various structural performance functions and implement them in the FE solver, review academic papers in shape optimization, and expand the FE toolbox to enable full differentiability. The work will be benchmarked, tested, and documented for publication or internal communication. Prospective candidates can join Autodesk’s talent community to stay informed about new opportunities and company news. • Make the in-house finite element solver fully differentiable. • Extend shape parameterization from single components to multi-component assemblies. • Develop a gradient-based optimization scheme for mechanical assemblies. • Conduct original research on shape sensitivity and structural optimization. • Derive and implement sensitivity equations for various structural performance functions. • Review academic literature in the shape optimization domain. • Enhance and benchmark the FE toolbox for full differentiability. • Document research outcomes for publications or internal communication.

Key Responsibilities

• ▸differentiable fe
• ▸shape parameterization
• ▸gradient optimization
• ▸sensitivity derivation
• ▸fe benchmarking
• ▸research publication

What You Bring

The Simulation, Optimization, and Systems (SOS) Group at Autodesk Research is seeking a dedicated and skilled research intern for Summer 2026. The position is located in the Toronto office at the MaRS Discovery District. The role involves making the in‑house finite element (FE) solver fully differentiable, extending shape parameterization to assemblies of interconnected components, and developing a gradient‑based optimization scheme for mechanical assemblies. Candidates must be currently pursuing a PhD in Mathematics, Physics, Engineering sciences, or a related discipline. They need strong knowledge of adjoint methods and gradient‑based constrained optimization, as well as solid experience with Python and/or C++ and scientific libraries. Experience developing shape optimization algorithms is required, and familiarity with variational PDE methods, free‑form deformation, or reduced‑order modeling is a strong plus. Proven ability to document scientific findings is also essential. • Pursuing a PhD in Mathematics, Physics, Engineering sciences, or a related field. • Strong knowledge of adjoint methods and constrained gradient-based optimization. • Proficient in Python and/or C++ with scientific libraries. • Experience developing shape optimization algorithms. • Familiarity with variational PDE methods, free‑form deformation, or reduced‑order modeling (plus). • Demonstrated ability to document scientific findings.

Requirements

• ▸phd
• ▸adjoint methods
• ▸gradient optimization
• ▸python
• ▸c++
• ▸shape optimization

Benefits

The 2026 Canada Intern Program runs for 16 weeks (May 4 – August 21) and is fully paid. Interns will work on meaningful projects, receive mentorship from industry leaders, and participate in tech talks and development activities. The program follows Autodesk’s Flexible Workplace approach, offering office, remote, or hybrid work options. Autodesk’s culture emphasizes innovation, sustainability, and belonging. Employees help create everything from green buildings to smart factories and blockbuster movies, guided by a culture that shapes how they work, collaborate, and engage with customers. The company values diversity, offering a competitive compensation package that reflects experience, education, and location. • 16‑week paid internship (May 4 – Aug 21) with mentorship, tech talks, and professional development. • Flexible workplace options (office, remote, hybrid).

Work Environment

Hybrid