Senior Cloud Infrastructure and Network Operations Solutions Architect
Core
Architect and operate compute, storage, and scale-up fabrics for large GPU estates, serving as a trusted advisor and hands-on technical leader for NVIDIA's Cloud Partner model.
Role type
Senior IC Network and Fabric Operations Solutions Architect
Builds
Production-stable InfiniBand/Ethernet fabrics for AI/HPC systems using open-source-first stacks (Cumulus Linux, SONiC, Spectrum-X, DOCA/DPU, Kubernetes)
Domain
Data Centre Networking / AI/HPC Infrastructure
Deliverable
production ML models | infrastructure
Required skills
Data centre network architecture, RDMA fabrics (InfiniBand, RoCE), NVLink/NVSwitch partition management, Cumulus Linux/SONiC, Linux internals, Python/Bash scripting, GitOps, observability stacks (Prometheus/Grafana)
Preferred skills
Kubernetes networking in GPU clusters, DOCA/DPU infrastructure services, fabric health telemetry (DCGM/XID), large-scale training traffic tuning, multi-tenant network isolation
Technologies
InfiniBand, RoCE, Ethernet, NVLink, NVSwitch, Cumulus Linux, SONiC, Spectrum-X, DOCA, Kubernetes, Prometheus, Grafana, Ansible, Terraform
Responsibilities
Manage Day 2 fabric operations including NVLink partition management and firmware lifecycle; Support Day 1 fabric validation and acceptance; Drive fleet reliability through telemetry analysis and fault remediation; Provide consultative guidance and troubleshooting across the fabric stack; Produce runbooks for partner teams; Lead architectural reviews and present to stakeholders
Seniority
Senior, hands-on IC
