Nvidia · posted · checked
Director, Software Engineering – Cloud Compute and Infrastructure
AWSAzureLinux$320k – $489kOn-site · Santa ClaraManager
Who can apply
On-site · Santa Clara
Location: US, CA, Santa Clara
NVIDIA is seeking an engineering director to lead the software teams behind our bare-metal datacenters and cloud compute infrastructure. The organization supports hundreds of megawatts of datacenter capacity already online, with more capacity coming. You will build the software that makes this growing physical infrastructure available as reliable, secure, and efficient compute services for NVIDIA engineering.
Our team provides NVIDIA’s continuous integration (CI) infrastructure: the environment where hardware, firmware, drivers, networking, and system software are integrated and tested on their path to production. You will enable engineering teams to bring up preproduction systems, reproduce failures, validate changes, and move platforms toward production readiness. The fleet spans multiple hardware generations and maturity levels: x86 and Arm servers, GPUs, DPUs, high-speed networking, storage, and rack-scale, liquid-cooled systems. Platforms such as Grace Blackwell and Vera Rubin illustrate the breadth of compute and interconnect technology involved. This role combines hands-on systems judgment with leadership of the teams making that diversity manageable at scale.
What you’ll be doing:
Lead software engineering teams responsible for cloud compute, bare-metal fleet management, and CI infrastructure, owning architecture, implementation, deployment, and operations.
Set the technical direction for compute control planes, resource provisioning, topology-aware placement, reservations, and capacity management across physical servers, virtual machines, and containers.
Automate the bare-metal lifecycle: hardware discovery and inventory, server bring-up, imaging, firmware and driver configuration, out-of-band management, health checks, reprovisioning, and recovery.
Build CI services that allocate the right hardware, provision repeatable test environments, record hardware and software configurations, collect diagnostics, and restore systems to a known state. Shorten the path from a hardware or software change to actionable validation results.
Keep engineering services dependable while supporting evolving preproduction hardware and software. Define service-level objectives, isolate failures, improve observability, and turn incidents and recurring test-infrastructure failures into engineering fixes.
Measure and improve hardware onboarding time, provisioning speed, CI queue time, productive fleet utilization, recovery time, and cost efficiency as capacity expands.