Senior Software Engineer, Bare Metal Infrastructure – DGX Cloud

NVIDIA•Santa Clara, CA

20h•$152,000 - $287,500

About The Position

NVIDIA DGX Cloud’s Bare Metal Infrastructure team builds and operates software that turns newly racked hardware into reliable, validated production capacity for demanding AI workloads. The engineer will contribute across the bare-metal lifecycle, including new platform introduction, firmware qualification, automated provisioning, validation, remediation, and handoff to production. The team uses AI agents and LLM tools extensively in software development and seeks someone who is interested in advancing their use.

Requirements

BS or MS in Computer Science, Computer Engineering, or equivalent experience.
5+ years building production infrastructure software, automation, or services, including substantial hands-on work with bare-metal server systems.
Strong programming skills in Rust, Go, or Python, with a track record of designing, testing, shipping, and maintaining production code and willingness to work in Rust regularly.
Deep knowledge of enterprise server hardware from multiple vendors, including Dell, Supermicro, Lenovo, and HPE, as well as BMCs, BIOS/UEFI, and firmware management.
Direct experience with Redfish and IPMI for provisioning, health inspection, power management, or fault diagnosis, and with network boot such as PXE/iPXE, DHCP, or HTTP boot.
Strong Linux knowledge, including the boot process, kernel and driver interactions, hardware discovery, and low-level failure analysis.
Solid datacenter networking fundamentals and ability to trace connectivity problems across host, management, and infrastructure networks.
Demonstrated, regular use of AI agents and LLM tools in software development, strong interest in improving, and judgment to validate their output.

Nice To Haves

Production experience writing Rust.
Experience with NVIDIA DGX or NVL72 systems, NVLink, InfiniBand, Spectrum-X, or BlueField DPUs.
Background in hardware qualification, rack-scale bring-up, firmware certification, or new product introduction, including safe, scalable workflows for fleet firmware upgrades, configuration enforcement, hardware replacement, and recovery.
Experience with Kubernetes, GitOps/Argo CD, or fleet-management platforms.
Examples of AI-driven tooling or workflows built that measurably changed how you or your team ships software.

Responsibilities

Design, build, and maintain production services for bare-metal discovery, provisioning, hardware validation, firmware upgrades, auditing, and repair across large GPU fleets.
Develop integrations with Redfish and vendor BMC interfaces to manage server state, collect inventory and health data, enforce configuration, and drive recovery while handling vendor differences in code.
Contribute to and review code in the Rust codebase, and build supporting tools and services in Go and Python.
Turn recurring hardware, firmware, and connectivity failures into automated detection, validation, and repair workflows.
Define and implement automated validation and handoff criteria for new platforms, firmware releases, and rack-scale deployments.
Debug complex failures across hardware, firmware, Linux, networking, and provisioning services, and incorporate findings into the software.
Partner with server and component vendors and internal platform, networking, datacenter, and production engineering teams to resolve cross-boundary issues.
Use AI agents and LLM tools daily to write, review, and test code and investigate failures, and help shape how the team builds with AI.