GPU Compute Systems Engineer (Kubernetes/Slurm)

Media Stream AI Limited; Dundee; Hybrid; GBP 54,000 - 62,000; Full time

Job summary

Media Stream AI Limited in Dundee is seeking an experienced HPC engineer to own the GPU compute estate, including servers, fabric, orchestration, and cluster health.

You will maintain GPU nodes, run burn-in tests, manage RoCE/InfiniBand, and lead Kubernetes and Slurm orchestration for GPUaaS deployments in a Linux-led environment.

The role requires strong Linux proficiency, GPU/HPC background, and networking know-how to keep systems performing at scale.

Qualifications

• Strong Linux expertise and experience with GPU clusters or HPC environments.
• Solid networking skills and understanding of high-performance fabric.
• Experience with orchestration tools like Kubernetes or Slurm.

Responsibilities

• Maintain B300 and H200 nodes and run burn-in and RMA.
• Manage the RoCE / InfiniBand fabric.
• Run Kubernetes and Slurm orchestration.
• Support GPUaaS client deployments.

Skills

Linux

GPU/HPC experience

Networking

Kubernetes

Slurm

Tools

Kubernetes

Slurm

Job description

Media Stream AI Limited in Dundee is seeking an experienced HPC engineer to own the GPU compute estate, including servers, fabric, orchestration, and cluster health.

You will maintain GPU nodes, run burn-in tests, manage RoCE/InfiniBand, and lead Kubernetes and Slurm orchestration for GPUaaS deployments in a Linux-led environment.

The role requires strong Linux proficiency, GPU/HPC background, and networking know-how to keep systems performing at scale.