Hardware Engineer

Infrastructure operations · shared across sites

Reports to: Director, Hardware Engineering

Location: Pleasanton, CA (hybrid) or assigned site; travel up to 25%

Department: Infrastructure & DC Operations / Systems Engineering

The Hardware Engineer owns hardware lifecycle for GPU and supporting infrastructure assets, including fleet health monitoring, RMA workflows, firmware management, and long-range capacity planning. The role is the technical owner of the physical compute platform.

• Monitor GPU and server health including thermal, error rates, and component failures

• Drive the RMA process with vendors (NVIDIA, Supermicro, HPE, and others) end-to-end

• Manage firmware, BIOS, and BMC upgrade campaigns across the fleet

• Develop hardware burn-in and acceptance test procedures, including NCCL and stress tests

• Investigate hardware failures and produce vendor-grade root cause analyses

• Maintain hardware inventory, asset records, and CMDB accuracy

• Drive capacity planning across compute, storage, and networking

• Coordinate with Procurement on spare parts strategy and stocking levels

• Author hardware engineering runbooks and operational procedures

• Support new platform bring-up, qualification, and reference architecture validation

• 5+ years in hardware engineering, systems engineering, or data center engineering

• Deep knowledge of x86 server architecture, GPU systems, and modern storage

• Hands-on experience with NVIDIA HGX, DGX, or hyperscale-class systems

• Strong Linux fundamentals and scripting skills (Python, Bash)

• Bachelor's degree in computer science, electrical engineering, or related field

• Experience with NVIDIA Mission Control, Base Command Manager, or Bright Cluster Manager

• Familiarity with IPMI, Redfish, and vendor management interfaces

• Knowledge of liquid cooling and high-density power architectures

• Experience operating fleets of 1,000+ GPUs