Senior Bare-Metal GPU Infra SRE – Kubernetes & DPUs | Equity

NVIDIA Corporation; Santa Clara (CA); Vor Ort; USD 152.000 - 288.000; Vollzeit

Benefits dieser Stelle

Equity

Benefits

Zusammenfassung

NVIDIA DGX Cloud engineers build and operate large-scale GPU infrastructure for AI workloads. You will contribute to production systems and automation across bare-metal provisioning, firmware upgrades, and cluster lifecycle management.

The role emphasizes hands-on experience with BMC/Redfish, Linux, and Kubernetes, plus on-call incident handling. NVIDIA values ownership and collaboration across hardware, networking, and data-center teams.

Qualifikationen

• 5+ years building software for or operating production infrastructure.
• Strong Go or Python skills with production automation.
• Direct experience with BMC and Redfish in provisioning or diagnostics.
• Practical experience with NVIDIA GPU hardware and DPUs.

Aufgaben

• Build automation for bare-metal provisioning, hardware validation, firmware upgrades, repair, and cluster lifecycle management.
• Develop tools that interact with BMC and Redfish interfaces to monitor hardware health and manage server state.
• Handle NVIDIA NVL72 systems and BlueField DPUs across cloud and on-prem environments.
• Diagnose failures across servers, DPUs, GPUs, CPUs, networking, Linux, and Kubernetes; automate detection and repair.
• Define validation and handoff criteria for safe, consistent capacity entry into production.
• Participate in on-call duties, incident response, root-cause analysis, and durable follow-ups.

Kenntnisse

Go

Python

Kubernetes

Linux

Observability

Ausbildung

BS/MS in Computer Science

Tools

BMC

Redfish

Argo CD

GitOps

Jobbeschreibung

NVIDIA DGX Cloud engineers build and operate large-scale GPU infrastructure for AI workloads. You will contribute to production systems and automation across bare-metal provisioning, firmware upgrades, and cluster lifecycle management.

The role emphasizes hands-on experience with BMC/Redfish, Linux, and Kubernetes, plus on-call incident handling. NVIDIA values ownership and collaboration across hardware, networking, and data-center teams.