Remote GPU/HPC SRE for Large-Scale AI Inference

Saturn Cloud

United Arab Emirates

Remote

AED 360,000 - 600,000

Full time

7 days ago

Be an early applicant

Application generator

Don’t send a generic resume — generate a resume and cover letter tailored to this exact role.

Get past ATS filters

Job summary

Saturn Cloud is seeking a GPU/HPC-focused Site Reliability Engineer for remote work across the MENA region. The role centers on GPU infrastructure, NVIDIA software, and Kubernetes-based workloads, with a focus on diagnosing and resolving production issues in large-scale GPU deployments.

You will work with NVIDIA drivers, NVML, and high-performance networking to ensure reliability and performance of inference workloads.

Qualifications

Hands-on GPU datacenter administration and driver troubleshooting.
Experience with Kubernetes GPU workloads and NVIDIA Operator.
Strong debugging skills for GPU, driver, and hardware issues.
Familiarity with CUDA, NVML, and nvidia-smi for diagnostics.

Responsibilities

Diagnose and resolve production issues on large-scale GPU inference infra.
Troubleshoot NVIDIA GPUs, drivers, CUDA compatibility, and GPU runtimes.
Investigate GPU health issues, Xid errors, and hardware/driver failures.
Analyze PCIe topology, NUMA, NVLink/NVSwitch, and GPU placement.
Troubleshoot multi-GPU and multi-node workloads in a Kubernetes context.
Work with GPU network fabric and distributed communication failures.
Differentiate between app, GPU, and infrastructure root causes; provide evidence.
Operate through Kubernetes-based GPU workloads and NVIDIA Operator plugins.