Network Engineer (Supercomputer Infrastructure)

Location: Memphis, Tennessee, or Southaven, Mississippi, United States. On-site infrastructure work includes job walks, physical network installation, working at heights and in tight quarters, travel between campuses of up to 20%, and after-hours on-call support.

Company

xAI develops AI systems intended to understand the universe and support humanity's pursuit of knowledge.

What you will do

• Design, deploy, and operate highly available, low-latency, high-bandwidth networks for GPU training and inference, storage, site operations, and OT/controls.
• Maintain supercomputer data-center and campus networks while collaborating with infrastructure, compute, storage, SiteOps, enterprise, facilities, vendors, and contractors.
• Evaluate, procure, and deploy data-center switches, NICs, firewalls, optical multiplexers, and other 400G/800G networking hardware.
• Develop network automation using configuration analysis, linting, validation, GitOps, and Infrastructure as Code frameworks.
• Troubleshoot cluster-impacting issues, provide networking support during bring-up and expansion, document RCAs, and participate in on-call operations.
• Build monitoring and telemetry for fabric health, congestion, packet loss, and NCCL/collective performance, and maintain network architecture and cabling documentation.

Requirements

• Bachelor's degree in computer science, computer engineering, or another STEM discipline with 3+ years of professional network engineering experience, or 5+ years of experience in lieu of a degree.
• Hands-on experience designing, deploying, supporting, and troubleshooting Layer 2 and Layer 3 networks in latency-sensitive, industrial, or data-center environments.
• Experience with multiple network vendors and with GitOps and Infrastructure as Code frameworks as both a user and contributor.
• Knowledge of network standards, routing, congestion control, redundancy, segmentation, QoS, multicast, and monitoring.
• Ability to pass applicable background checks, work in tight quarters and at heights, lift 30 lbs, drive with an active valid driver's license, and work extended hours or weekends.
• Ability to provide 24x7 emergency on-call support and participate in an after-hours rotation is required.

Nice to have

• Experience with Cisco, Arista, Juniper, or NVIDIA Spectrum-X data-center switches.
• RoCEv2 Ethernet AI/HPC fabric experience; InfiniBand and NCCL familiarity are a plus.
• Experience with WDM, single-mode and multimode fiber plants, OTDR, and acceptance testing.
• Proficiency in Bash, PowerShell, or Python and automation tools such as Terraform and Ansible.
• CCNA or CCNP certification, Linux and Windows administration experience, or experience with real-time, industrial control, energy, aerospace, defense, or other high-reliability systems.

Culture & Benefits

• Work in a small, highly motivated organization focused on engineering excellence and curiosity.
• Operate in a flat structure where employees are expected to contribute hands-on to the mission.
• Collaborate directly with internal teams, customers, vendors, and contractors.
• Provide flexible after-hours coverage aligned with cluster build-out and operational schedules.