Compute Electrical Engineer, Robotics

About the Role

We're looking for an Electrical Engineer to develop the AI inference compute systems that power OpenAI's humanoid robots. This role combines compute architecture, high-speed digital design, onboard power electronics, and compact HDI PCB development.

We're looking for a self-driven engineer who takes ownership from concept through deployment. Above all, you should enjoy hands-on engineering: bringing up boards, characterizing interfaces, and tracking challenging failures to their root cause.

In this role, you will:

• Own compute hardware architecture and development, from component selection and schematic design through PCB layout, validation, and deployment.
• Design and integrate high-speed interfaces—including PCIe, DDR, Ethernet, MIPI, USB, and DisplayPort—balancing bandwidth, connectivity, power, and signal integrity.
• Develop onboard DC-to-DC conversion and power distribution to meet efficiency, transient-response, sequencing, and power-integrity requirements.
• Define HDI PCB stackups, dielectric materials, via structures, and routing strategies that balance electrical performance, manufacturability, reliability, and cost.
• Lead bench bring-up and validation, including test planning, high-speed interface characterization, power measurements, and debugging through verified resolution.
• Partner with thermal, mechanical, EMC, and software teams to integrate compute hardware into tightly constrained robotic systems.

You might thrive in this role if you:

• Have 5+ years of experience developing compute hardware or similarly complex high-speed electronics, with demonstrated ownership of complete designs.
• Bring strong hands-on measurement and debugging skills, including practical experience with oscilloscopes, probing techniques, and interface validation.
• Understand high-speed digital interfaces deeply enough to make architectural tradeoffs and diagnose issues across hardware and software boundaries.
• Have strong signal-integrity and power-integrity fundamentals, a solid foundation in power electronics, and practical experience designing complex HDI PCBs.
• Can use a Linux shell, write simple shell scripts, connect over SSH, and perform basic Ethernet and network troubleshooting.
• Take initiative, turn ambiguous requirements into concrete engineering plans, and follow through to deliver working systems.

Additional, preferred qualifications:

• Experience designing around AI inference SoCs or accelerators, including their power, memory, and communication architectures.
• Familiarity with how AI inference workloads map to silicon capabilities, including compute throughput (TOPS), DRAM bandwidth and capacity, and how workload requirements and DRAM technology choices affect achievable performance and performance per watt.
• Experience bringing compute hardware into production and resolving fabrication, assembly, or test issues with manufacturing partners.