NVIDIA H800 AI Supercomputing Support Engineer

OneAsia Network Limited; Hong Kong; On-site; HKD 600,000 - 900,000; Full time

Job summary

OneAsia Network Limited in Hong Kong is seeking a skilled NVIDIA H800 Support Engineer to manage enterprise-scale AI supercomputing environments. You will provide deep technical troubleshooting for DGX SuperPOD deployments across multi-node clusters.

You bring strong Linux system administration, high-performance networking and experience with NVIDIA GPU tools, InfiniBand, Slurm or Kubernetes, and Python/Bash scripting to automate operations.

Qualifications

• Bachelor's or Master's degree in computer science, computer engineering, or other engineering fields or equivalent experience.
• 3+ years in providing in-depth system support and debugging experience.
• System level expertise of CPU/GPU server architecture, NICs, Linux, system software and kernel drivers.
• Proven hands-on experience in Linux troubleshooting with good problem identification, resolution and solving skills.
• Experience with NVIDIA GPU infrastructure tools (DCGM, nvidia-smi, MIG, NVSwitch diagnostics)
• Hands-on experience with InfiniBand protocols, RDMA, and RoCEv2 architectures.
• Experience with cluster orchestrators like Slurm or Kubernetes.
• Proficiency in Bash and Python scripting for system monitoring and automation.

Responsibilities

• Diagnose and resolve multi-node hardware and software failures across clusters.
• Guide customer discussions on network design, compute/storage and support bring up of server/network/cluster deployments.
• Troubleshoot and fine-tune ultra-high-bandwidth fabrics to maximize compute efficiency.
• Perform deep-dive post-mortems on systemic bugs, node drops and cluster regressions.
• Partner with vendor support teams to patch software bugs and systemvulnerability.

Skills

Linux troubleshooting

NVIDIA DGX

NVIDIA GPU tools

InfiniBand RDMA

Slurm Kubernetes

Bash scripting

Python scripting

Cluster management

Networking design

Education

Bachelor/Master in CS/CE

Tools

DCGM

nvidia-smi

NCCL tuning

Docker

Kubernetes

Job description

OneAsia Network Limited in Hong Kong is seeking a skilled NVIDIA H800 Support Engineer to manage enterprise-scale AI supercomputing environments. You will provide deep technical troubleshooting for DGX SuperPOD deployments across multi-node clusters.

You bring strong Linux system administration, high-performance networking and experience with NVIDIA GPU tools, InfiniBand, Slurm or Kubernetes, and Python/Bash scripting to automate operations.