Principal Architect, System Software - Orbital Data Center

Details

Space-1 is NVIDIA's first Orbital Data Center (ODC) module, a Vera Rubin-class compute platform engineered for low-Earth orbit missions. This role owns the end-to-end system software architecture for Space-1 and successor orbital platforms, spanning applications, libraries, data center software, BMC and BIOS firmware, manageability and telemetry, the host OS, GPU and CPU drivers, and CUDA. The goal is to deliver a production-ready inference platform that operates reliably in radiation, thermal-cycling, and remote-operations environments in low-Earth orbit.

Responsibilities

• Own the system architecture for inference stacks and other applications, ensuring resilience to faults occurring in space.
• Co-architect with the orbital hardware system architecture team to define interfaces, partitioning, and trade-offs across silicon, boards, firmware, operating systems, and AI workload layers for five-year LEO missions.
• Own end-to-end system software architecture for Space-1 and successor Orbital Data Center modules, including the data center stack, BMC firmware, BIOS, host OS, GPU and CPU drivers, CUDA, DCGM, manageability, and telemetry.
• Define manageability architecture for an unreachable, autonomous data center, including remote bring-up, in-orbit firmware updates, dual-module redundancy, fault containment, recovery from SEU/SEFI events, and telemetry for fleets ranging from tens to millions of nodes.
• Architect radiation-tolerant system software behaviors, including ECC handling, memory scrubbing, latch-up mitigation, deterministic recovery, and graceful degradation through five years and up to approximately 8,000 thermal cycles.
• Drive Redfish, MCTP, PLDM, and constellation-level management protocols across BMC, BIOS, and host software.
• Define the minimum BMC feature set, pin budget, boot architecture, rugged M.2 or VPX-class options, and dual-module redundancy strategy in partnership with platform and mechanical engineering.
• Partner with cloud and constellation customers, including SpaceX, Blue Origin, Starcloud, Planet, and Cowboy Space, to translate mission requirements into platform software architecture.
• Drive reliability and optimization in the system software architecture, including operation through eclipse periods and idle-power retention strategies.
• Work with the bring-up team to resolve issues from first silicon through first launch, and own the quality, reliability, and telemetry performance of system software delivered with each ODC module.

Requirements

• 15+ years of relevant experience in server or platform system software spanning compute libraries, BMC firmware, BIOS, host operating systems, drivers, and manageability.
• BS, MS, or PhD in electrical engineering, computer science, or a related field, or equivalent experience.
• Experience building AI infrastructure and systems in space, with a proven record of architecting and delivering platform software for large-scale data centers or mission-critical embedded systems.
• Strong knowledge of server architecture, data center manageability, and full-stack integration of firmware with operating systems and accelerator software.
• Hands-on experience with data center health management workflows, telemetry, and fault management at scale.
• Understanding of USB, SMBus/I2C, and PCIe hardware management interfaces.
• Proficiency with Redfish, MCTP, and PLDM.
• Strong, demonstrable skills in C/C++ and Python.
• Experience programming and debugging server platforms, including pre-silicon and platform bring-up environments.
• Experience with source code management tools such as Git or Perforce and project management tools such as Jira.
• Excellent written and oral communication skills, teamwork, initiative, and commitment to producing high-quality work.
• Ability to find creative solutions to complex problems and remain hands-on with coding.

Preferred Qualifications

• Experience architecting platform software for space, aerospace, defense, or other radiation-, thermal-, and vibration-constrained environments, including SEU/SEFI mitigation, ECC strategy, TID/SEE qualification, and radiation-hardened design partitioning.
• Experience with autonomous, remote, or unreachable data center operations, including in-orbit or in-field firmware updates, dual-module redundancy, and recovery without physical access.
• Experience with x86 or ARM system architectures, including Grace/Vera, and the NVIDIA AI software stack, including CUDA, DCGM, DOCA/OFED, GPU drivers, and DGX OS.
• Familiarity with NSA PHIPs security, post-quantum networking, and aerospace standards such as VPX, MIL-STD shock/vibration, and NASA EEE-INST-002.
• Proven technical leadership of large, complex programs involving 50+ engineers across firmware, operating system, driver, and AI stack teams.
• Ability to review hardware schematics and PCB layouts for debugging, design verification, and collaboration with hardware engineers.

Benefits

• Base salary range of USD 272,000 to USD 431,250, determined by location, experience, and the pay of employees in similar positions.
• Eligibility for equity and benefits.
• NVIDIA is an equal opportunity employer committed to an inclusive work environment.