Senior Network Engineer, AI/HPC Networking
Job Description
Job Description
Senior Network Engineer, AI/HPC Networking
Location: San Jose, CA · On-site · Full-time
Base salary range: $180,000 – $260,000 per year
About ARC Cloud
ARC Cloud Inc is an AI infrastructure company building and operating GPU cloud and data center capacity for AI training and inference workloads. Our teams design and run the network, storage, and cluster platforms that keep large GPU clusters performing reliably for customers.
About the role
Design and operate the high-performance networks that connect large GPU clusters. You'll work across InfiniBand and/or RoCE fabrics, Ethernet, RDMA, and high-speed switching to help distributed AI workloads move data with low latency and predictable performance. This is a hands-on engineering role with room to influence architecture, automation, and operational standards.
What you'll do
- Design, deploy, and operate scalable network fabrics for GPU and HPC environments, including leaf-spine and cluster-specific topologies.
- Implement and troubleshoot InfiniBand and/or RoCE, RDMA, congestion control, routing, switching, and high-speed links.
- Analyze packet loss, latency, congestion, link health, and NCCL or similar workload-level performance signals.
- Develop network provisioning, configuration validation, telemetry, alerting, and remediation automation.
- Partner with compute and storage engineers to isolate bottlenecks across the full AI infrastructure stack.
- Manage network hardware lifecycle, including switches, NICs, optics, cabling standards, spares, and maintenance windows.
- Maintain accurate topology diagrams, configuration standards, runbooks, and change records; participate in incident response.
What we're looking for
- 5+ years of network engineering experience in data center, cloud, telecom, HPC, or AI infrastructure environments.
- Strong Ethernet and data center networking fundamentals, including TCP/IP, routing, switching, VLANs, and BGP.
- Hands-on experience with InfiniBand and/or RoCE and a practical understanding of RDMA-based workloads.
- Experience troubleshooting high-throughput, low-latency networks and interpreting network performance data.
- Experience with network automation or scripting using Python, Ansible, or comparable tools.
- Clear technical communication and sound change-management practices in production environments.
Helpful, but not required
- 400G/800G networks, NVIDIA/Mellanox ConnectX adapters, NVIDIA Spectrum-X, or SONiC.
- EVPN/VXLAN, OSPF, Nornir, and automated network testing/validation.
- NCCL performance analysis or experience supporting large-scale distributed AI training clusters.
- CCNP, JNCIP, or equivalent demonstrated expertise.
How success will be measured
Stable, measurable network performance; reduced time to diagnose fabric issues; repeatable deployment practices; and clear operational documentation.
Compensation
The base salary range for this role in San Jose, California is $180,000 – $260,000 per year. The range shown is a good-faith estimate of the base salary we reasonably expect to pay for this role in this location. Actual compensation will depend on experience, skills, qualifications, and internal equity. This role may also be eligible for additional compensation and benefits, to be confirmed at the offer stage.
How to apply
Please submit a resume and a brief note describing the experience most relevant to this role. Candidates do not need to match every preferred qualification to apply.
ARC Cloud Inc is an equal opportunity employer. We consider all qualified applicants without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, veteran status, or any other status protected by applicable law.
