SoC Engineer at OpenAI in San Francisco

ABOUT THE TEAM OpenAI’s Hardware organization develops custom silicon and systems for advanced AI workloads. Our first-party accelerator team works across chip architecture, design, software, and systems to build the computing platforms that power frontier models. We are a small, dynamic team that values technical breadth, hands-on engineering, and creative approaches to building complex hardware. ABOUT THE ROLE We’re looking for a SoC Engineer to help architect, design, and integrate OpenAI’s first-party AI accelerators. You will build the processor, interconnect, I/O, and platform infrastructure that connects our accelerator designs into complete, functional systems. This role combines architectural judgment with hands-on implementation. You’ll work closely with design, architecture, verification, and software teams throughout development, from defining subsystem requirements and integrating IP to pre-silicon validation, silicon bring-up, and post-silicon debug. You should enjoy stretching across domains, rolling up your sleeves, and doing the work needed to make a complex system function end to end. IN THIS ROLE, YOU WILL: - Define and implement SoC architectures spanning embedded processors, AMBA interconnects, peripheral subsystems, PCI Express, Ethernet and die-to-die interfaces. - Develop the microarchitecture and RTL design of subsystems supporting I/O and SoC…