DevOps Engineer for HEP GPU Benchmarks (IT-CE-LCG-2026-168-GRAP)
Core
Develop, operate, and support the WLCG HEPScore benchmarking software stack to measure and compare heterogeneous computing resources (CPUs, GPUs, Cloud, HPC) for the High Energy Physics community.
Role type
DevOps Engineer (GPU Benchmarking)
Builds
Next-generation benchmarking software stack, containerized HEP workloads, and performance analysis services.
Domain
High Energy Physics / High Performance Computing / Grid Infrastructure
Required skills
Python, Docker, Apptainer/Singularity, GitLab CI/CD, Puppet, Linux administration, Bash, GPU profiling (CUDA/ROCm), OpenSearch, Logstash, Kibana, Grafana, Apache Spark
Preferred skills
NVIDIA Container Toolkit, ActiveMQ, system metrics tracking
Technologies
Python, Bash, Git, Docker, Apptainer/Singularity, GitLab CI/CD, Puppet, OpenSearch, Logstash, Kibana, Grafana, Apache Spark, ActiveMQ, CUDA, ROCm
Responsibilities
Develop and package HEP benchmark software; build and maintain container images via CI/CD; maintain benchmarking infrastructure and data ingestion services; analyze grid-wide benchmark data for performance scaling; provide technical support and documentation.
Seniority
Graduate / Applied Engineer (2-6 years post-grad experience or PhD with <3 years post-grad)
