Voltai
Computational Scientist
Palo Alto Office · Full-time · Posted Feb 19, 2026
- Education
- PhD in computational physics, applied mathematics, computational engineering, or a closely related field.
About the job
Develops and validates distributed GPU PDE solvers and neural-operator surrogates for simulating fields on complex 3D integrated-circuit geometries.
What you'll do
- Develop and optimize MPI+CUDA PDE solvers and numerical methods for multi-node GPU clusters.
- Build neural-operator surrogates and simulation pipelines to generate training data.
- Validate solvers and learned models against analytical solutions, benchmarks, and each other.
What they ask for
- PhD in computational physics, applied mathematics, computational engineering, or a closely related field.
- Deep expertise in numerical PDE methods, such as FEM, FVM, or BEM.
- Strong C++ and CUDA skills, plus multi-node HPC experience with MPI.
- Expertise in sparse linear algebra and hands-on neural-operator experience, with knowledge of AI for Science methodology.
Nice to have
- Experience with HYPRE, PETSc, and Trilinos; familiarity with multi-node GPU clusters, NCCL, CUDA-aware MPI, or NVLink topologies.
- Published work in neural operators, physics-informed ML, or scientific HPC; IC design domain knowledge.
Tools
MPI, CUDA, C++, FEM, FVM, BEM, FNO, DeepONet
AI focus
Build and train neural operators as PDE solver surrogates; design simulation datasets, address out-of-distribution generalization, and ensure physical consistency.
Model-written summary · the original posting is authoritative