AI GPU Compute Network Architect

Tencent; Singapore; On-site; SGD 180,000 - 240,000; Full time

Job summary

Tencent seeks an experienced AI Infrastructure Network Engineer / Architect to plan, deliver, and operate large-scale GPU cluster networks for internal AI teams. You will bridge internal AI demand with external vendors across regions, validating designs and ensuring production readiness.

The role requires strong data center networking, GPU interconnect expertise, and cross‑team coordination to deliver scalable, reliable AI compute clusters.

Qualifications

• Bachelor's degree or higher in a relevant technical field.
• 5+ years in data center networking, cloud infra, or large-scale delivery.
• Solid AI infrastructure understanding for distributed training workloads.
• Familiar with InfiniBand and RoCEv2 in data center contexts.
• Experience with vendor reviews, design validation, and production readiness.

Responsibilities

• Translate business/workload requirements into GPU cluster network specs.
• Review and validate GPU cluster network architecture and plans.
• Engage with external vendors and system integrators for delivery.
• Monitor delivery progress, identify risks, coordinate corrective actions.
• Define acceptance criteria and support performance validation.
• Coordinate handover to internal teams with docs and procedures.

Skills

Data center networking

GPU cluster networking

InfiniBand networking

RoCEv2

Spine-Leaf topology

Vendor management

Network troubleshooting

Capacity planning

Cross-team collab

Linux networking

Education

Bachelor's degree in CS/EE/IT

Tools

InfiniBand

RoCEv2

VXLAN

Spine-Leaf

Monitoring tools

TCP/IP

DNS

Firewall rules

Job description

Tencent seeks an experienced AI Infrastructure Network Engineer / Architect to plan, deliver, and operate large-scale GPU cluster networks for internal AI teams. You will bridge internal AI demand with external vendors across regions, validating designs and ensuring production readiness.

The role requires strong data center networking, GPU interconnect expertise, and cross‑team coordination to deliver scalable, reliable AI compute clusters.