Compute Hardware Engineer: GPU & Server Qualification

Linuxcareers; Austin, Northern (TX, KY); Hybrid; USD 120,000 - 180,000; Full time

Job summary

FluidStack is building civilization-scale compute infrastructure for AI. The Compute Engineer, R&D role focuses on qualifying next-generation GPUs, servers, and racks before deployment to the fleet, ensuring they meet performance, power, and thermal standards.

The candidate will build benchmarks, perform stress tests, and write reports to standardize fleet configurations. Linux and hardware vendors collaboration are part of the daily workflow, with a focus on reproducible methods and automation.

Qualifications

• Hands-on experience qualifying, validating, or debugging server or GPU systems.
• Rigorous benchmarking methodology with reproducible methods and controlled variables.
• Ability to root-cause hardware failures across silicon, firmware, and system levels.
• Scripting and automation skills to drive work programmatically rather than via checklists.
• Experience working with hardware vendors on bug fixes and patches.

Responsibilities

• Qualify and validate GPUs, servers, and racks before fleet deployment.
• Build and run benchmark suites to characterize performance, power, and thermal behavior.
• Conduct stress testing, fault injection, and endurance runs to uncover weaknesses.
• Root-cause hardware failures across silicon, firmware, and system levels.
• Write qualification reports documenting configurations, firmware, and issues.

Skills

Hardware qualification

Benchmarking methodology

Root-cause analysis

Scripting & automation

Vendor collaboration

Tools

GPU/Server validation tooling

Linux systems depth

Job description

FluidStack is building civilization-scale compute infrastructure for AI. The Compute Engineer, R&D role focuses on qualifying next-generation GPUs, servers, and racks before deployment to the fleet, ensuring they meet performance, power, and thermal standards.

The candidate will build benchmarks, perform stress tests, and write reports to standardize fleet configurations. Linux and hardware vendors collaboration are part of the daily workflow, with a focus on reproducible methods and automation.