SLURM HPC Architect​/Administrator

SLURM HPC Architect / Administrator

Location:

Remote (Canada, U.S., or Europe Preferred)
Company:
Cylix Applied Intelligence

Employment Type:

Full-Time or Contract

Cylix Applied Intelligence is seeking an experienced SLURM HPC Architect / Administrator to design, deploy, and operate high-performance computing (HPC) clusters supporting AI training, large‑scale inference, scientific computing, and enterprise workloads.

This role will focus on building and managing enterprise‑grade HPC environments powered by GPU and CPU compute clusters, leveraging SLURM as the core workload orchestration and resource scheduling platform.

You will work closely with AI engineers, infrastructure teams, and enterprise clients to deliver scalable, reliable, and high‑performance compute environments across on‑premise, hybrid, and cloud platforms.

• Design and implement SLURM‑based HPC cluster architectures
• Architect scalable CPU and GPU compute environments
• Define cluster topology including compute, storage, login, and management nodes
• Design high‑availability SLURM controller configurations
• Implement cluster segmentation, partitioning, and resource allocation strategies

• Install, configure, and manage SLURM workload manager environments
• Configure SLURM partitions, queues, QoS policies, and scheduling policies
• Manage job scheduling optimization and fair‑share policies
• Implement accounting, usage tracking, and reporting systems
• Maintain SLURM cluster health, stability, and performance

• Configure GPU scheduling and allocation policies
• Support GPU resource management including:
  • NVIDIA A100, H100, L40, and similar accelerator platforms
  • MIG partitioning and GPU isolation
  • Multi‑tenant GPU resource allocation
• Optimize cluster performance for AI training and inference workloads

• Automate cluster deployment and configuration using:
  • Ansible, Terraform, or similar tools
  • Shell scripting and Python
• Implement monitoring, alerting, and performance tracking systems
• Support cluster lifecycle management, upgrades, and expansion

• Integrate HPC clusters with high-performance storage systems including:
  • NFS
  • Lustre
  • BeeGFS
  • GPFS / Spectrum Scale
• Optimize I/O performance and storage architecture

• Support enterprise and research users with job scheduling and optimization
• Troubleshoot job failures and performance issues
• Assist engineering teams in optimizing workloads for HPC environments

• 3+ years experience administering HPC clusters
• Strong experience with SLURM workload manager
• Strong Linux system administration experience (Ubuntu, Rocky Linux, RHEL, or similar)
• Experience with HPC cluster architecture and deployment
• Experience with shell scripting and automation
• Experience with:
  • Cluster resource management
  • Multi-node distributed computing environments
  • SSH, networking, and Linux system internals

• Experience managing GPU-based HPC clusters
• Experience supporting AI / ML workloads
• Experience with NVIDIA GPU platforms and drivers
• Experience with:
  • CUDA environments
  • NVIDIA MIG configuration
  • GPU scheduling optimization
• Experience with configuration management tools:
  • Ansible
  • Terraform
  • Puppet or Chef
• Experience with monitoring tools such as:
  • Prometheus
  • Grafana
  • Node exporter
  • SLURM accounting tools

• Experience with large-scale enterprise or cloud HPC environments
• Experience deploying HPC environments in cloud platforms such as:
  • AWS
  • Azure
  • Private cloud environments
• Experience with containerized HPC workloads:
  • Docker
  • Singularity / Apptainer
• Experience integrating SLURM with Kubernetes or hybrid orchestration systems