HPC Operations Engineer: High-Performance Cluster Lead

Evergreen Statistical Trading; Bellevue (KY); On-site; USD 180,000 - 220,000; Full time

Job summary

Evergreen is a proprietary trading firm based in Bellevue, Washington. As an HPC Operations Engineer, you will own the day-to-day operation of our research clusters across multiple datacenters and the scheduled workloads our research and trading rely on.

You will work on a modern stack without legacy infrastructure and take on major projects across software and hardware as our HPC footprint expands, collaborating with a hands-on team.

Qualifications

• Linux systems administration on Red Hat-family distros (RHEL, Rocky, AlmaLinux)
• Python for automation and tooling
• Git and versioned configuration management experience
• Prometheus/Grafana monitoring and alerting experience
• Slurm batch scheduler experience including queue config and resource management
• Cluster provisioning tools such as Warewulf or xCAT
• ZFS and parallel filesystems like GPFS or Lustre
• InfiniBand networking experience
• Ability to own critical processes and resolve issues outside business hours
• Hands-on hardware coordination in datacenters
• Ability to diagnose complex production problems under time pressure
• Growth-oriented and collaborative mindset
• Passion for high-performance infrastructure with minimal drama

Responsibilities

• Own day-to-day operation of research clusters across multiple datacenters and scheduled workloads
• Work on a modern stack unencumbered by legacy infrastructure and tackle major projects across software and hardware as HPC footprint grows
• Collaborate with teams to scale and improve HPC performance and reliability

Skills

Linux systems administration

Python scripting

Ownership of critical processes

Team collaboration

Tools

Git

Configuration management

Prometheus

Grafana

Slurm

Warewulf

xCAT

ZFS

GPFS

Lustre

InfiniBand

Job description

Evergreen is a proprietary trading firm based in Bellevue, Washington. As an HPC Operations Engineer, you will own the day-to-day operation of our research clusters across multiple datacenters and the scheduled workloads our research and trading rely on.

You will work on a modern stack without legacy infrastructure and take on major projects across software and hardware as our HPC footprint expands, collaborating with a hands-on team.