Power AI workloads with AMD Instinct™ GPUs. Scale models faster with high-performance, dedicated cloud compute.
About the role
We are looking for a Hardware Diagnostics Engineer to run burn-in, triage what fails, work servers out-of-band, and own RMAs end to end. If you like hardware that misbehaves in ways that take real work to explain, this is a good seat. Before any GPU server carries a customer workload, it has to prove it works — under load, at temperature, for hours. When it doesn't, somebody has to figure out why, get replacement hardware in, and send the failed part back to the vendor. Found on 1752vc Careers, the job board for startup and VC roles.
What they're looking for
- 3–6 years in datacenter operations, systems administration, hardware support, or infrastructure engineering
- Hands-on experience with enterprise server hardware: component replacement, POST and boot failures, and reading hardware behavior at the rack
- Practical experience with BMCs and out-of-band management: IPMI, Redfish, iDRAC, iLO, or equivalent
- Strong Linux troubleshooting: boot process, driver and device issues, and diagnostic tools such as {{dmesg}}, {{lspci}}, {{ipmitool}}, and SMART
- Comfort reading sensor data, event logs, and thermal and power telemetry well enough to tell a real failure from noise
- Working scripting ability in Bash or Python — enough to automate a repetitive task and read someone else's tooling
More about this role
Our mission is simple: deliver seamless, secure, reliable, and resilient AI compute at scale. We've built a versatile cloud platform that eliminates infrastructure barriers, empowering builders to focus on innovation instead of fighting their stack. Because breakthrough AI should move at the speed of ideas, not infrastructure.
We are looking for a Hardware Diagnostics Engineer to run burn-in, triage what fails, work servers out-of-band, and own RMAs end to end. If you like hardware that misbehaves in ways that take real work to explain, this is a good seat.
Before any GPU server carries a customer workload, it has to prove it works — under load, at temperature, for hours. When it doesn't, somebody has to figure out why, get replacement hardware in, and send the failed part back to the vendor.
Run server and GPU burn-in and stress testing, interpret the results, and decide whether hardware is production-ready
Triage failures across GPUs, memory, drives, NICs, PSUs, and cabling: reproduce the failure, isolate the faulty component, and document what proved it
Work servers out-of-band through IPMI and Redfish for power control, boot configuration, BIOS settings, and sensor and event...
Browse similar: Startup jobs