Onsite Data Center Technician (GPU Infrastructure)

Quick Facts

• Role: Onsite Data Center Technician (GPU Infrastructure)

• Openings: 3

• Location: Downtown Los Angeles, CA (on-site at a customer data center)

Description

You will be hands-on onsite for a multi-thousand-GPU cluster, fixing hardware failures on the floor while using Linux and networking knowledge to determine why issues occurred. The job covers GPU tray, NVMe, memory, PSUs, NICs, and optics troubleshooting using tools such as BMC/Redfish, DCGM, and vendor utilities, alongside RMAs, change control, and root-cause reviews.

Responsibilities

• Fix hardware failures: GPU trays, NVMe, memory, PSUs, NICs, and optics using BMC/Redfish, DCGM, and vendor tools

• Fix and validate links: inspect/replace optics; test fiber and copper; trace flapping ports with the network team

• Monitor the floor: respond to rack-level power, temperature, and coolant alarms; escalate facility issues early

• Run RMAs and manage spares: pull failed parts, ship returns, re-validate replacements, keep NetBox accurate

• Follow change control: disruptive work (power-cycling, component swaps, firmware) under approved procedures

• Write and document: clear tickets, MOPs, and handover notes; contribute to root-cause reviews

• On-call participation: hands-on escalations with remote hands covering routine overnight work

Requirements

• 3+ years of hands-on work with servers, cabling, and networks in data centers or HPC (5+ years for the senior opening)

• Working Linux knowledge: logs, services, and basic scripting in Bash or Python

• Experience with BMC/IPMI/Redfish and a methodical approach to hardware fault diagnosis

• Fiber and optics handling: cleaning, inspection, DOM readings

• Ability to maintain careful written records of what was found and what was done

• Ability to lift up to 50 lb and work in a loud, cold data hall

• Live within 60 minutes of downtown Los Angeles

Benefits

• Base salary: $105,000–$145,000 depending on level, plus equity and an on-call stipend

• Full-time, on-site role in Los Angeles with Monday–Friday shifts from 7 a.m. to 7 p.m.

• On-call rotation with on-site response within 60 minutes when paged