HPC Data Center Ops Engineer

Sfcompute; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Benefits offered by this job

Generous equity grant

401(k) matching 4%

Medical/dental/vision insurance

Unlimited PTO

Holidays

Daily lunch

Book budget

Job summary

San Francisco Compute operates GPU data centers and supercomputers for lease. This hands-on role covers racking, commissioning, and maintaining HGX/DGX-class GPU systems, high-density compute infrastructure, and associated cabling and cooling.

You will handle procurement, BOMs, vendor quotes, and RMAs, keep audit-ready asset records, and develop runbooks while mentoring a small, growing team. This position is on-site in San Francisco and requires working with energized hardware in

Qualifications

• 5+ years of hands-on data center operations with hardware and cabling responsibility.
• Expertise with high-density fiber optics and transceivers; diagnose faulty connectors.
• Experience racking, cabling and commissioning equipment in energized environments.
• Procurement knowledge: BOMs, vendor quotes, POs, receiving, warranties and RMAs.
• Proficiency with TCP/IP, DHCP, DNS, VLANs, IPMI/iDRAC/iLO/Redfish, and Linux CLI.

Responsibilities

• Rack, commission, and maintain HGX/DGX-class GPU systems and high-density compute infrastructure.
• Replace failed components, manage fiber optics (transceivers, end-face inspection) and liquid cooling.
• Manage data center procurement including BOMs, quotes, POs, receiving, warranties and RMAs.
• Maintain accurate asset records and inventory across sites with audit-ready docs.
• Establish runbooks and mentor team members in a small, growing group.

Skills

Data center operations

High-density fiber optics

Hardware racking & cabling

Network basics (TCP/IP, DNS, DHCP, VL​

Procurement & BOMs

Tools

IPMI/iDRAC/iLO/Redfish

NetBox

Linux CLI

Job description

San Francisco Compute operates GPU data centers and supercomputers for lease. This role is hands-on data center operations, managing the infrastructure that keeps high-performance compute clusters running reliably. You will handle everything from racking and commissioning GPU systems to managing failed components, fiber optics, liquid cooling, and procurement.

What You'll Do

• Rack, commission, and maintain HGX/DGX-class GPU systems and high-density compute infrastructure
• Replace failed server components (FRUs), manage fiber optics including transceiver swaps and end-face inspection, and maintain liquid cooling systems
• Manage data center procurement including BOMs, vendor quotes, purchase orders, receiving, warranties, and RMAs
• Maintain accurate asset records and inventory across multiple sites, with audit-ready documentation
• Establish operational standards, develop runbooks, and mentor team members as part of a small, growing team

What You Need

• 5+ years of hands-on data center operations, infrastructure support, or systems administration with direct hardware and cabling responsibility
• Expertise with high-density fiber optics: breakouts, end-face inspection, cleaning, and transceiver swaps; ability to diagnose faulty connectors and optics
• Experience racking, cabling, and commissioning equipment safely in energized environments; proficiency replacing server FRUs
• Practical procurement knowledge: BOMs, vendor quotes, purchase orders, receiving, warranty management, and RMA processes
• Proficiency with TCP/IP, DHCP, DNS, VLANs, out-of-band management protocols (IPMI, iDRAC, iLO, Redfish), and Linux command line

Nice to Have

• Experience in HPC, GPU, or high-density compute environments including HGX/DGX platforms and rack-scale designs
• Familiarity with InfiniBand or 400G+ Ethernet networking
• Experience with liquid cooling systems: CDUs, manifolds, and direct-to-chip cold plates
• Colocation facility experience and understanding of tenant/operator relationships
• NetBox or similar DCIM, and procurement or ERP systems; scripting for automation (Bash, Python, or PowerShell)
• Enterprise wireless, MDM, and identity provider experience; certifications such as CompTIA Server+, CCNA, or NVIDIA credentials

Competitive salary + generous equity grant, 401(k) matching up to 4%, medical/dental/vision insurance, unlimited paid time off, 10+ observed holidays, parental leave, daily lunch, unlimited office book budget