Senior Site Reliability Engineer - Kubernetes Platform NVIDIA · US, CA, Santa Clara · Site Reliability In listing 2d ago

NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years. It's a unique legacy of innovation fueled by great technology—and amazing people. The GPU, our invention, serves as the visual cortex of modern computers and sits at the heart of our products and services. Every one of those GPUs starts life inside our silicon design environment, and that environment runs on the platform this team builds.

We're looking for a Senior SRE to own and scale the Kubernetes platform that assists NVIDIA's chip design (EDA) workloads. This is not a generic container-platform role. You'll be running Kubernetes at the edges of what it was developed for: orchestrating enormous stateful jobs, license-constrained EDA tooling, latency-sensitive access to high-performance shared storage, and tenants whose tapeout schedules do not move.

The clusters you run sit alongside some of the largest EDA compute farms in the industry, spanning multiple on-prem datacenters and multiple public clouds. When these clusters degrade, chip schedules feel it - which is exactly what makes the reliability work exciting.

• Own Kubernetes clusters end-to-end - plan, deliver, upgrade, day-2 operations, and decommission - across on-prem datacenters and managed cloud Kubernetes (e.g. EKS, GKE, RKE, Red Hat Open Shift, etc.).
• Define and implement the multi-cluster lifecycle strategy: version skew policy, patch cadence, node image pipelines, controlled rollouts, and safe upgrade mechanics for clusters that can't simply be drained on demand.
• Treat on-prem and CSP Kubernetes as one platform with consistent APIs, policy, and observability - not as two disconnected worlds.
• Drive everything through Infrastructure-as-Code, Git Ops, and declarative config; manual cluster changes should be an anomaly worth investigating.
• Design for failure: explicit fault domains, blast-radius containment between clusters and tenants, progressive delivery, and disciplined change control.
• Build end-to-end observability (metrics, logs, traces, events) and push toward data-driven and auto-healing operations that materially reduce human intervention.
• Participate in on-call, lead incident response, drive root cause analysis, and publish high-quality RCAs that actually change the system.
• Partner with EDA engineering, storage, and network teams to land changes cleanly in a fast-moving environment.
• Engage and partner with data center operations to help forecast, design and implement the accurate on-prem and cloud deployments to meet architectural needs.
• Interface directly with internal EDA tenants/engineering teams to design and implement operational improvements to increase productivity and provide faster and more reliable outcomes as well as partner with the infrastructure services team to help architect the supporting capabilities for intelligent workload routing, observability, dynamic scaling and resource governance.

• Bachelor's degree in Computer Science or related technical field (or equivalent experience) with 8+ years building and supporting critical production services.
• Deep, hands-on Kubernetes operations experience - you've run it, not just consumed it. Comfortable debugging etcd, kubelet, CNI, CSI, scheduler behavior, API server load, container runtime, and control-plane degradation under real production pressure.
• Production experience with CSP-managed Kubernetes — including its upgrade model, networking and IAM integration, node group lifecycle, and where the managed abstraction stops helping you.
• Production experience operating self-managed Kubernetes on-premises across multiple datacenters — cluster bring-up, bare-metal and GPU node lifecycle, multi-site topology, and the…