Doctoral Student in High-Performance Computing

About the role

Four-year fully funded PhD at KTH on high-performance computing, focused on scalable algorithms, parallel runtimes, and code-generation for modern accelerators (GPUs and AI chips). The host group has access to LUMI, Dardel, and partner supercomputing facilities. Deadline 20 June 2026.

Responsibilities

• Design parallel algorithms for scientific simulation or ML training workloads.
• Build and optimise runtime systems, schedulers, or compilers for heterogeneous hardware.
• Run scaling experiments on national HPC facilities and publish results.
• Take 60 ECTS of doctoral coursework over the four-year programme.

Requirements

• Master’s degree in computer science, applied mathematics, computational physics, or a related field.
• Solid C++ and a parallel-programming stack (MPI, OpenMP, CUDA, SYCL, or HIP).
• Strong grasp of computer architecture and numerical methods.
• Fluent English.

Nice to have

• Experience with at least one supercomputing platform.
• Background in compiler design, performance analysis, or numerical linear algebra.
• A first-author paper or open-source contribution.

How to apply

Open the apply link and submit CV, transcripts, research statement, Master’s thesis, and two reference contacts via the KTH recruitment system before 20 June 2026.