We are looking for motivated candidates to join a vibrant and collaborative team of scientists and engineers in the Advanced Manufacturing & Semiconductor Division (AMS) at the Institute of High Performance Computing (IHPC), A*STAR. The successful candidate will contribute to research and development in AI for Science and scientific machine learning, focusing on the development of Physics Foundation Models (PFM) for PDE-governed physical systems relevant to advanced manufacturing and semiconductor technologies.
This research explores new computational paradigms that integrate neural operators, graph-based representations of numerical solvers, and transformer architectures to enable structurally grounded machine learning models for complex physical systems such as electromagnetic fields, multiphysics processes, and device-scale simulations.
You will work on R&D projects spanning fundamental methodology development and application-driven research, with opportunities to collaborate with interdisciplinary teams and industrial partners in semiconductor and advanced manufacturing domains.
The key scope of work includes:
- Developing new scientific machine learning methods for modelling PDE-governed physical systems.
- Designing and advancing neural operator architectures for physics-based modelling and simulation.
- Exploring graph-based representations of numerical solvers (e.g., finite-difference or finite-element methods) for integrating physical inductive biases into machine learning models.
- Developing hybrid computational architectures that combine neural operators, graph neural networks, and transformer-based models.
- Building physics-informed machine learning frameworks that incorporate governing equations, boundary conditions, and numerical solvers.
- Developing data-driven surrogate models and simulation acceleration methods for manufacturing and semiconductor applications.
- Publishing research outcomes in leading journals and conferences in AI for Science, scientific machine learning, and computational physics.
- Collaborating with internal research teams, industry partners, and affiliated institutes on interdisciplinary R&D projects.
Job Requirements:
- PhD degree in Computer Science, Applied Mathematics, Computational Physics, Electrical Engineering, Mechanical Engineering, or related disciplines with at least 5 years of relevant experience.
- Strong background in scientific machine learning, numerical simulation, or computational physics.
- Solid understanding of PDE-based physical modelling and numerical methods.
- Experience in machine learning and deep learning, particularly in areas such as neural operators, graph neural networks, or physics-informed learning.
- Strong programming skills in Python, PyTorch, JAX, or similar machine learning frameworks.
- Experience with high-performance computing or large-scale simulations is an advantage.
- Strong analytical and problem-solving skills, with the ability to work both independently and collaboratively.
- We particularly welcome early-career researchers who are passionate about advancing AI-driven scientific discovery.
The above eligibility criteria are not exhaustive. A*STAR may include additional selection criteria based on its prevailing recruitment policies. These policies may be amended from time to time without notice. We regret that only shortlisted candidates will be notified.