Skip to content
← All roles
SF Compute

HPC Data Center Ops Engineer

A market for GPU time: buyers reserve H100 to B300 clusters for a chosen start date and length, as VMs, bare metal or managed Slurm, and resell any hours they…

San Francisco, CAFull-time

Checked 20 Nov

What you’d do

  • Keep SF Compute's GPU clusters running by hand: rack and commission systems, swap failed parts, chase bad optics.
  • Manage fibre, transceivers, liquid cooling and accurate asset records.
  • Buy the hardware too: parts lists, quotes, lead times, spares, warranties and returns.
  • Spend about 80% on data centre work and buying, and 20% on office IT.
  • Visit local sites often and remote sites now and then, with some after-hours windows.
  • Help set standards and write runbooks for a small team.

What they look for

  • 5+ years in data centre operations or systems administration, with real hands-on hardware and cabling.
  • Knows fibre well: inspection, cleaning and transceiver swaps.
  • Has racked, cabled and commissioned equipment and replaced server parts safely.
  • Has handled purchasing, receiving, warranties and returns.
  • Keeps inventory records that would survive an audit.
  • Networking basics and out-of-band management (IPMI, iDRAC, iLO, Redfish), plus Linux.
  • Based in the San Francisco Bay Area, and able to lift up to 50 lbs.

Helpful: HGX or DGX-class GPU systems; InfiniBand or 400G Ethernet; liquid cooling; NetBox; scripting in Bash or Python.

Perks: 401(k) match up to 4%; health, dental and vision cover for you and dependants; unlimited paid time off; parental leave; lunch every day.

How SF Compute hires, and every role open there →
Know the company before you applyFull company page →
funding, as the company states it
Series A

SF Compute hires through MenFem. How, and what that means for our review, is on the company page.

Other roles that fit