Principal Software Engineer - Compute Infrastructure

NVIDIA•Santa Clara, CA

1d•Hybrid

About The Position

NVIDIA has been reinventing computer graphics, PC gaming, and accelerated computing for 30 years. It is a unique legacy of innovation that’s fueled by great technology and amazing people. Today, we’re tapping into the unlimited potential of AI to define the next era of computing. An era in which our GPU acts as the brains of computers, generative AI, robots, and self-driving cars that can understand the world. Doing what’s never been done before takes vision, innovation, and the world’s best talent. As an NVIDIAN, you’ll be immersed in a diverse, supportive environment where everyone is inspired to do their best work. We are seeking a highly skilled Principal Software Engineer to join our dynamic team. Our company is at the forefront of technological innovation, and we are dedicated to driving efficiency, defining platform architecture, and optimizing the performance of our infrastructure both on-prem and in the cloud. You will lead the architectural vision for a massive global platform and spearhead the operationalization of our internal frontier-class AI inference systems. Join us in this exciting endeavor!

Requirements

Bachelor’s degree in Engineering, Computer Science, Mathematics, or related field, or equivalent experience.
15+ years of proven experience in compute platform engineering, site reliability, or systems architecture with a heavy focus on automation at massive scale.
Deep expertise in Kubernetes architecture and designing/deploying virtualization architectures, specifically operating VMs inside K8s (KubeVirt, OpenShift).
In-depth knowledge of hardware technologies (GPUs, high-speed backplane networking) with a track record of mitigating hardware-level failures, silent data corruption, and anomalies in large-scale environments.
Experience running large global environments spanning bare metal, virtualized infrastructure, and cloud with a unified GitOps posture (ArgoCD or similar).
Proficiency in programming languages such as Go and/or Python, alongside expert-level infrastructure-as-code development (Terraform, Config Management).
Strong leadership skills with the ability to influence technical direction across highly autonomous teams without relying on top-down mandates.

Nice To Haves

Hands-on experience managing bleeding-edge, pre-release hardware in production environments.
Deep understanding of advanced storage migrations and protocols (NFSv4, NVMe/TCP, Hyperconverged storage).
Solid understanding of microservices architecture and seamless multi-cloud deployment strategies (AWS, GCP).
Proven track record of building "Day 2" operational maturity (self-service, advanced auto-remediation, strict SLAs) from the ground up on existing foundations.

Responsibilities

Define Platform Architecture: Lead initiatives to architect and transform our global enterprise compute platform—running thousands of nodes and tens of thousands of VMs and containers via OpenShift and KubeVirt—by defining service tiers, SLAs, and automated cluster lifecycles.
Operationalize Frontier AI Infrastructure: Build the operational foundation for our internal AI inference platform scaling to frontier-class models. You will develop automated remediation pipelines, hardware watchdogs, and telemetry for pre-release, rack-scale GPU systems (including Blackwell and upcoming architectures).
Drive Strategic Capacity & Scale: Collect and review system data for capacity planning to navigate extreme hardware supply constraints. Develop proactive strategies, including public cloud bursting, hardware dogfooding, and evaluating alternative compute architectures (e.g., ARM).
Build the "Paved Road": Collaborate with highly autonomous NVIDIA engineering teams to drive cultural adoption of standard platforms. You will design compelling self-service architectures, APIs, and Terraform/OpenTofu providers that teams want to use.
Lead Complex Migrations: Evaluate existing application architectures and drive the fraught but critical migration of massive legacy workloads—including large-scale, long-running VDI environments—into modern Kubernetes orchestration.