Senior Network Engineer

1. Key Responsibilities1.1 AI Cluster Network Architecture Design & Deployment

• Design and deploy AI cluster network architectures based on  NVIDIA Spectrum-X Ethernet  and  NVIDIA Quantum InfiniBand  technologies.
• Build  Leaf-Spine and Clos network architectures  covering GPU compute networks, storage networks, management networks, and cross-site connectivity.
• Develop standard reference architectures for  single-tenant and multi-tenant GPUaaS environments , including network segmentation, routing domains, and business zones.
• Evaluate and select network hardware based on performance, cost, reliability, scalability, and security requirements.

1.2 RDMA, RoCEv2 & InfiniBand Operations and Optimization

• Configure, operate, troubleshoot, and optimize  RoCEv2 and InfiniBand lossless networks .
• Perform in-depth RDMA performance tuning for distributed AI/ML workloads, including  NCCL, MPI, and inter-GPU traffic .
• Manage and optimize  congestion control, ECN, PFC, QoS, and adaptive routing .
• Troubleshoot communication bottlenecks caused by switches, NICs, DPUs, optical transceivers, firmware, and the Linux networking stack.

1.3 Cluster Integration & Performance Validation

• Work closely with compute, storage, and data center teams to integrate and commission GPUs, NICs, DPUs, storage nodes, and related infrastructure.
• Conduct network and cluster performance testing using tools such as  ib_write_bw, iperf, NCCL test suites , and other benchmarking tools.
• Establish performance baselines covering latency, bandwidth, packet loss, CRC errors, and other critical network metrics.
• Support customer  POCs, benchmark testing, technical validation, and production deployment .

1.4 Multi-Tenant Network Services

• Implement tenant isolation using  VLAN, VRF, VXLAN/EVPN , and related data center networking technologies.
• Design and deploy customer connectivity solutions including dedicated circuits, cloud interconnects,  IPsec/VPN , and other secure access services.
• Operate network services including firewalls, load balancers, DNS, DHCP, and out-of-band management.
• Develop standardized deployment and onboarding templates to improve tenant provisioning efficiency.

1.5 Automation, Monitoring & Operations

• Develop network automation frameworks using  Python/Go, Ansible, CI/CD, and Infrastructure as Code (IaC) .
• Implement  Zero-Touch Provisioning (ZTP) , configuration validation, automated deployment, and version rollback capabilities.
• Build comprehensive telemetry and monitoring covering port status, optical power levels, congestion, RDMA metrics, and other critical infrastructure indicators.
• Develop and maintain operational runbooks, escalation procedures, incident response processes, and  Root Cause Analysis (RCA)  templates.
• Support  24×7 production operations  and participate in on-call and emergency incident response.

1.6 Project Delivery & Technical Documentation

• Participate in data center planning, including rack layout, network cabling, IP addressing, hardware BOMs, and infrastructure design.
• Produce  High-Level Design (HLD), Low-Level Design (LLD), Method of Procedure (MOP), test plans, test reports , and other technical documentation.
• Provide technical support for major production incidents and participate in project deployments across the  Asia-Pacific region .
• Coordinate with data center operators, hardware vendors, customers, and internal technical teams to ensure successful project delivery.

2. Requirements2.1 Essential Requirements

• Bachelor’s degree or above in  Computer Science, Information Technology, Electronics, Telecommunications, or a related discipline .
• Minimum  5 years of experience  in large-scale data center network design, deployment, and troubleshooting.
• Minimum  3 years of hands-on experience  with  InfiniBand, RoCEv2, and RDMA .
• Proven experience supporting  AI/GPU clusters, High-Performance Computing (HPC), or large-scale distributed computing environments .

2.2 Required Technical Skills

• Strong knowledge of  BGP, ECMP, VXLAN/EVPN, QoS, high availability , and modern data center networking technologies.
• Hands-on experience with  NVIDIA/Mellanox  networking products, including:
• NVIDIA Spectrum switches
• NVIDIA ConnectX NICs
• NVIDIA BlueField DPUs
• Familiarity with networking equipment from major vendors such as  Arista, Cisco, and Juniper .
• Strong Linux networking troubleshooting skills using tools such as  tcpdump, ethtool , and related utilities.
• Experience with network monitoring and observability platforms such as  Prometheus, Grafana, and ELK .

2.3 Preferred / Additional Qualifications

• Hands-on experience deploying  NVIDIA DGX, HGX, GB-series AI systems, Spectrum-X, or Quantum InfiniBand .
• Knowledge of  NCCL, GPUDirect RDMA, and large-scale AI/ML workload traffic patterns , including MoE workloads.
• Familiarity with  Kubernetes networking, Slurm, parallel file systems, NVMe-oF , and related AI/HPC technologies.
• Experience with  Python/Go, Ansible, CI/CD , and network automation.
• NVIDIA, Arista, or other relevant vendor certifications are an advantage.

3. Performance Indicators & Additional Requirements3.1 Key Performance Indicators

• Deliver network projects on schedule with complete and accurate cabling, configuration, testing, and acceptance documentation.
• Ensure  RDMA and NCCL performance  meets defined customer and production requirements.
• Respond rapidly to network incidents and accurately identify root causes, with complete RCA documentation and corrective action plans.
• Continuously improve network automation, operational efficiency, tenant security, and network isolation.

3.2 Additional Requirements

• Willingness to travel frequently across  Asia-Pacific  countries.
• Able to support scheduled night-time maintenance, emergency troubleshooting, and on-call operations when required.
• Comfortable working in a fast-paced project environment with multiple stakeholders.
• Strong communication and coordination skills when working with data center operators, hardware vendors, enterprise customers, and internal engineering teams.

4. Job Summary

The  AI Cluster Network Engineer  is a key technical role responsible for designing, deploying, optimizing, and operating high-performance networks for  AI/GPU computing and GPUaaS environments .

The role is deeply focused on NVIDIA’s advanced networking technologies, including  Spectrum-X Ethernet, Quantum InfiniBand, RoCEv2, RDMA, and high-performance cluster networking . The successful candidate will combine strong traditional data center networking expertise with hands-on experience in AI/HPC networking and distributed workload optimization.

In addition to network architecture and performance tuning, the role covers  multi-tenant network isolation, automation, observability, production operations, customer POCs, and project delivery  across the Asia-Pacific region.

This position is ideal for a senior network engineer who wants to specialize in  AI infrastructure, GPU clusters, high-performance computing, and next-generation data center networking , with a strong focus on commercial, production-grade GPUaaS deployments.