Jens Glaser
Computational Scientist, Materials
Knoxville, Tennessee, United States
Summary
🤩
Rockstar🎓
Top SchoolJens Glaser is a computational scientist with 26 years of experience applying molecular dynamics, Monte Carlo methods, and high-performance computing to soft and biological materials. Based at Oak Ridge National Laboratory, he develops and optimizes codes for leadership-class supercomputers while supporting users in large-scale simulations. His work spans algorithm development and performance engineering, including enabling GPU support across CUDA and AMD HIP and improving cross-platform robustness in projects like the widely used hoomd-blue and BlazingSQL. Trained as a physicist (Ph.D., Leipzig), Jens combines deep academic research with pragmatic software and DevOps skills, quietly specializing in memory-mapped communication and build portability that keep scientific codes performant on diverse architectures.
26 years of coding experience
4 years of employment as a software developer
Ph.D., Physics, Ph.D., Physics at Leipzig University
Github Skills (19)
c-language10
python10
molecular-simulation10
ucx10
hip10
rapids10
molecular-dynamics-simulation10
gpu10
nonlinear-dynamics10
cuda10
cprogramming-language10
gpu-acceleration10
asyncio9
nvcc9
build-system9
Programming languages (7)
Github contributions (5)
glotzerlab/hoomd-blue
Aug 2019 - Jul 2022
Molecular dynamics and Monte Carlo soft matter simulation on GPUs.
Role in this project:
Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:39 reviews, 195 commits, 59 PRs in 2 years 11 months
Contributions summary:Jens primarily focused on enabling and optimizing the `hoomd-blue` molecular dynamics simulation for AMD GPUs. Their contributions involve modifying CMake files and code to support HIP (Heterogeneous-compute Interface for Portability), fixing compilation errors related to HIP, and re-enabling compilation with nvcc. They also made minor changes to the code and data structures to improve performance. These modifications likely involve changes to CUDA/HIP kernels, compiler flags, and library configurations to ensure the application functions correctly and efficiently.
pythonhard-particlesimulationmonte-carlodocker
BlazingDB/blazingsql
Jul 2020 - Sep 2021
BlazingSQL is a lightweight, GPU accelerated, SQL engine for Python. Built on RAPIDS cuDF.
Role in this project:Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:3 reviews, 15 commits, 2 PRs in 1 year 2 months
Contributions summary:Jens focused on enhancing the communication and infrastructure of the BlazingSQL engine. Their contributions include refactoring communication threads using asyncio and UCX, optimizing memory management with UCP memory mapping. Furthermore, they worked on build system improvements, including patches for activation scripts and disabling Java bindings for UCX builds, contributing to the portability and maintainability of the project on powerpc architecture. These changes indicate a focus on performance and cross-platform compatibility.
cudfgpu-accelerationpythondata-sciencegpu
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.