Associate Computational Plasma Scientist — HPC & ML

Princeton University; Princeton (NJ); On-site; USD 96,000 - 153,000; Full time

Job summary

The Princeton Plasma Physics Laboratory (PPPL) at Princeton University seeks an Associate Computational Scientist to advance computational modeling of low-temperature plasmas using HPC, novel algorithms, and machine learning to accelerate codes and build surrogate models.

Responsibilities include developing and maintaining the LTP-PIC software, collaborating with academic and industrial partners, and applying AI-driven methods to improve plasma modeling and simulation throughput.

Qualifications

• Ph.D. in Physics, Engineering or a related field with core training in low-temperature plasma physics and high-performance computing.
• Minimum 3 years of professional experience in an academic, scientific, or R&D environment is preferred.
• A proven track record of publishing original results in peer-reviewed journals.
• A proven track record of developing high-performance codes.

Responsibilities

• Ongoing development and maintenance of the LTP-PIC software and assisting users from academia and industry (50%).
• Defining and delivering on A.I. projects for low-temperature plasmas (20%).
• Modeling low-pressure discharges for industry partners (20%).
• Publishing scientific results and presenting at major international conferences (10%).

Skills

Python

NumPy

PyTorch

HPC programming

Machine learning

PIC algorithms

Education

Ph.D. in Physics/Engineering

Tools

LTP-PIC

MPI

OpenMP

OpenACC

Job description

The Princeton Plasma Physics Laboratory (PPPL) at Princeton University seeks an Associate Computational Scientist to advance computational modeling of low-temperature plasmas using HPC, novel algorithms, and machine learning to accelerate codes and build surrogate models.

Responsibilities include developing and maintaining the LTP-PIC software, collaborating with academic and industrial partners, and applying AI-driven methods to improve plasma modeling and simulation throughput.