Onboard Payload Processing HPC Engineer — Sunrise

Blue Origin; Denver (CO); Hybrid; USD 182,000 - 254,000; Full time

Job summary

Blue Origin LLC in Denver, CO, is seeking a Payload Processing Software Engineer to own the onboard processing runtime for high‑performance workloads. You will integrate accelerator frameworks, optimize GPU‑based compute and ensure fault tolerance across intermittent satellite contacts.

You’ll design robust stateful systems, define job submission APIs, and build a benchmarking suite to characterize latency, throughput, and power tradeoffs on reference hardware.

Qualifications

• BS in CS/CE or related field required.
• 6+ years experience in HPC, distributed systems, or accelerator software.
• Strong knowledge of GPU architectures, memory bandwidth, and parallel models.
• Experience with fault-tolerant distributed systems in intermittent connectivity.

Responsibilities

• Own the architecture and implementation of the on-board payload processing runtime for diverse, high-performance workloads.
• Integrate, configure, and optimize accelerator compute frameworks and libraries, including defining workload optimization strategies for our orbital hardware.
• Design and write the core system code from scratch for robust workload checkpointing and state persistence to handle interruptions.
• Define the API contract for job submission and results delivery, collaborating with the compute orchestration team.
• Develop and maintain a performance benchmark suite on reference hardware to characterize throughput, latency, and power tradeoffs.

Skills

HPC experience

Distributed systems

Accelerator computing

Fault tolerance

Parallel programming

GPU architecture

Education

Bachelor of Science in CS/CE or related field

Master’s or PhD in relevant discipline

Tools

Docker

Kubernetes

CUDA

Profiling tools

Job description

Blue Origin LLC in Denver, CO, is seeking a Payload Processing Software Engineer to own the onboard processing runtime for high‑performance workloads. You will integrate accelerator frameworks, optimize GPU‑based compute and ensure fault tolerance across intermittent satellite contacts.

You’ll design robust stateful systems, define job submission APIs, and build a benchmarking suite to characterize latency, throughput, and power tradeoffs on reference hardware.