GPU Network Architect for AI/HPC Clusters

CyberCoders; Santa Clara (CA); Hybrid; USD 200,000 - 250,000; Full time

Benefits offered by this job

Health Benefits

401k

Relocation assistance

Job summary

CyberCoders in Sunnyvale, CA is seeking a hands-on GPU Network Engineer to design, build and operate the high-speed fabrics that connect our GPU clusters. This hybrid role requires on-site collaboration and a strong focus on reliability and scalability.

You will deploy InfiniBand (NDR/XDR) and RoCEv2 interconnects, design CLOS/ECMP networks, and work with Kubernetes and bare-metal environments to move data efficiently at scale.

Qualifications

• 2+ years in data center or HPC networking with GPU/AI clusters.
• Hands-on GPU cluster networking design and operations.
• Deep experience with InfiniBand (NDR/XDR) and RoCEv2.
• Experience with Kubernetes and multi-host networking.
• Familiarity with Netbox, Netconf, and IaC tooling.

Responsibilities

• Design and deploy east-west network fabrics for GPU-to-GPU, rack-to-rack, and cluster-to-cluster communication.
• Build and operate InfiniBand (NDR/XDR) and RoCEv2 interconnect fabrics as the primary transport for GPU workloads.
• Implement CLOS/ECMP architectures for high-bandwidth, low-latency data movement across GPU clusters.

Skills

Data center networking

HPC networking

GPU cluster environments

InfiniBand (NDR/XDR)

RoCEv2 interconnects

Multi-host networking

Tools

Netbox

Netconf

Ansible

Terraform

KVM

Job description

CyberCoders in Sunnyvale, CA is seeking a hands-on GPU Network Engineer to design, build and operate the high-speed fabrics that connect our GPU clusters. This hybrid role requires on-site collaboration and a strong focus on reliability and scalability.

You will deploy InfiniBand (NDR/XDR) and RoCEv2 interconnects, design CLOS/ECMP networks, and work with Kubernetes and bare-metal environments to move data efficiently at scale.