Senior Cloud Infrastructure And Network Operations Solutions Architect

Nvidia; Sassari; On-site; EUR 120,000 - 180,000; Full time

Job summary

NVIDIA is seeking a Senior Cloud Infrastructure and Network Operations Architect to own Day 2 fabric operations for Cloud Partner estates, guiding InfiniBand, RoCE/Ethernet, and NVLink/NVSwitch across multi-tenant AI/HPC systems.

You will lead hardware bring-up validation, fleet reliability, and runbooks for partner teams, with hands-on tuning of NICs/DPUs, switch OS, and Kubernetes networking in a fast, open-source–first fabric ecosystem.

Qualifications

• 5+ years in data center networking and large-scale operations.
• Deep knowledge of InfiniBand, RoCE/Ethernet, and NVLink/NVSwitch.
• Experience with NVL72-class platforms and NICs/DPUs.
• Strong Linux (RedHat/Ubuntu) and switch OS skills.
• Automation skills: Python, Bash, Ansible, Terraform.
• Kubernetes networking in GPU clusters and multi-tenant isolation.

Responsibilities

• Own Day 2 fabric operations across NVIDIA Cloud Partner fleets (NVLink/NVSwitch partition management, maintenance-partition isolation, safe partition-change workflows)
• Manage switch software and firmware lifecycle including upgrades and rollout campaigns with minimal production disruption
• Support Day 1 fabric validation and acceptance (InfiniBand/UFM bring-up, Spectrum-X/RoCE config, burn-in against MTBI/goodput targets)
• Minimise handover time to production by reducing duplicated validation across hardware bring-up and partner operations
• Drive fleet-wide fabric reliability through telemetry, fault detection, remediation, and root-cause analysis of network-induced job failures
• Provide consultative guidance and hands-on troubleshooting across fabric stack (NICs/DPUs, switch OS, routing, Kubernetes networking) and lead knowledge transfer with runbooks for partner teams
• Serve as technical leader for assigned accounts and present to executive stakeholders
• Deep understanding of data center architectures and RDMA fabrics
• Hands-on experience with InfiniBand/UFM, Spectrum-X Ethernet, Cumulus Linux and/or SONiC
• Strong Linux knowledge and HPC/AI traffic patterns
• Automation and Observability skills: Python/Bash, IaC tools (Ansible, Terraform), GitOps, Grafana/Loki/Prometheus
• Proven ability to measure and improve MTBI and job goodput, and lead architectural reviews with executives
• Experience with Kubernetes networking in GPU clusters and multi-tenant network isolation concepts
• Strong consultative mindset with knowledge transfer and runbook creation
• Strong communication and presentation skills
• Cross-functional collaboration
• InfiniBand and UFM
• Spectrum-X Ethernet

Skills

Data center networking

Fabric engineering

NVLink/NVSwitch

InfiniBand

RoCE/Ethernet

RDMA fabrics

Linux

Automation

Python

Ansible

Terraform

Grafana

Prometheus

Kubernetes networking

Multi-tenant isolation

UFM

Spectrum-X

Cumulus Linux

SONiC

NICs/DPUs

ConnectX/BlueField

NVL72

Tools

Cumulus Linux

SONiC

ConnectX/BlueField

NICs/DPUs

Kubernetes

Job description

In this role you will own Day 2 fabric operations for NVIDIA’s Cloud Partner estates, guiding the reliability and performance of InfiniBand, RoCE/Ethernet, and NVLink/NVSwitch fabrics at scale. You’ll work with cross-functional teams and customers to architect, validate, and operate complex GPU-centric networks across multi-tenant environments. This is a hands-on leadership role in a fast-moving, open-source–first fabric ecosystem. You’ll drive operational excellence, incident response, and knowledge transfer to partner teams, shaping how large AI/HPC systems run reliably.

• Own Day 2 fabric operations across NVIDIA Cloud Partner fleets (NVLink/NVSwitch partition management, maintenance-partition isolation, safe partition-change workflows)
• Manage switch software and firmware lifecycle including upgrades and rollout campaigns with minimal production disruption
• Support Day 1 fabric validation and acceptance (InfiniBand/UFM bring-up, Spectrum-X/RoCE config, burn-in against MTBI/goodput targets)
• Minimise handover time to production by reducing duplicated validation across hardware bring-up and partner operations
• Drive fleet-wide fabric reliability through telemetry, fault detection, remediation, and root-cause analysis of network-induced job failures
• Provide consultative guidance and hands-on troubleshooting across fabric stack (NICs/DPUs, switch OS, routing, Kubernetes networking) and lead knowledge transfer with runbooks for partner teams
• Serve as technical leader for assigned accounts and present to executive stakeholders
• 5+ years in data center networking, fabric engineering, or large-scale network operations
• Deep understanding of data center architectures and RDMA fabrics (InfiniBand, RoCE/Ethernet) including topology, routing, congestion control, and QoS
• Hands-on experience with InfiniBand/UFM, Spectrum-X Ethernet, Cumulus Linux and/or SONiC, ConnectX/BlueField NICs/DPUs, NVLink/NVSwitch on NVL72-class platforms
• Strong Linux knowledge (RedHat/Ubuntu), switch OS internals, security, and HPC/AI traffic patterns
• Automation and Observability skills: Python/Bash, IaC tools (Ansible, Terraform), GitOps, Grafana/Loki/Prometheus
• Proven ability to measure and improve MTBI and job goodput, and to lead architectural reviews with executive stakeholders
• Experience with Kubernetes networking in GPU clusters and multi-tenant network isolation concepts
• Strong consultative mindset with demonstrated capacity to transfer knowledge and create runbooks for partner teams
• Consultative mindset
• Strong communication and presentation skills
• Cross-functional collaboration
• InfiniBand and UFM
• Spectrum-X Ethernet

Senior Cloud Infrastructure and Network Operations Solutions Architect sassari, sardegna, it