Senior HPC & AI Workload Orchestration Engineer

F. Hoffmann-La Roche AG; Madrid; Presencial; EUR 70.000 - 110.000; Jornada completa

Descripción de la vacante

F. Hoffmann-La Roche AG in Madrid seeks a Workload Orchestration Engineer within the ACE team to oversee and advance our workload orchestration stack for HPC and AI Factory platforms.

You will deploy, configure, and tune SLURM across clusters and run Run:ai to enable fractional GPU allocation and dynamic scheduling for multi-tenant workloads. You will define containerized execution using Singularity/Apptainer and Enroot, optimize QoS and queues, and collaborate with Observability to build

Formación

• 5+ years of systems engineering experience focusing on workload scheduling and multi-tenant cluster optimization.
• Deep familiarity with Enterprise Linux and distributed systems architectures.
• Expert-level proficiency in SLURM administration, partitions, accounting and plugins.
• Hands-on experience with Run:ai, Kubernetes, and GPU scheduling paradigms.
• Automation of scheduler configurations and telemetry collection.

Responsabilidades

• Orchestration Stack Deployment & Governance: design, deploy, and maintain SLURM across HPC cluster architectures.
• Deploy and manage Run:ai as the core AI orchestration layer for fractional GPU allocation.
• Integrate SLURM Slinky with Kubernetes where needed to bridge HPC and container-based workflows.
• Containerization & Workload Optimization: define best practices with Singularity/Apptainer/Enroot.
• Profile and tune queues, QoS, and fair-share policies for multi-tenant efficiency.
• Platform Reliability & Telemetry: implement monitoring, dashboards, and telemetry for scheduler performance.
• Troubleshoot distributed training, MPI/NCCL issues and driver compatibility.

Conocimientos

Workload scheduling

Resource management

Cluster optimization

Linux administration

Automation

Cross-team collaboration

Educación

Bachelor’s or advanced degree in Computer Science or related field

Herramientas

SLURM

Run:ai

Kubernetes

Singularity/Apptainer

Enroot

Descripción del empleo

F. Hoffmann-La Roche AG in Madrid seeks a Workload Orchestration Engineer within the ACE team to oversee and advance our workload orchestration stack for HPC and AI Factory platforms.

You will deploy, configure, and tune SLURM across clusters and run Run:ai to enable fractional GPU allocation and dynamic scheduling for multi-tenant workloads. You will define containerized execution using Singularity/Apptainer and Enroot, optimize QoS and queues, and collaborate with Observability to build