Senior HPC Networking Engineer for Space-Grade Interconnects

Blue Origin; Merritt Island (FL); On-site; USD 134,000 - 187,000; Full time

Benefits offered by this job

Medical, dental, vision

401(k) with company match

Paid time off

Education Support Program

Stock options

Job summary

Blue Origin seeks a hardware-leaning systems engineer to own the foundational hardware decisions for a high-speed interconnect fabric in a large-scale orbital data center. You will oversee from physical layer up to the software stack, ensuring resiliency to radiation and optimal performance for AI workloads.

You will collaborate with software teams to align fabric topology with model parallelism and participate in fabric bring-up and integration efforts.

Qualifications

• 8+ years in HPC networking, datacenter fabric design, or GPU cluster engineering.
• Experience with high-speed GPU-to-GPU interconnects and high-radix fabric switches.
• Hands-on knowledge of large-scale network topology and fabric management tools.
• Proficiency with collective communication libraries.
• U.S. citizen or eligible to work in the U.S.

Responsibilities

• Architect the hierarchical GPU interconnect topology for a large-scale accelerator cluster.
• Own fabric management software configuration, bring-up, and partitioning policies for multi-tenant workloads.
• Lead hardware qualification of fabric switches and other network components for the space environment.
• Tune communications libraries and define bandwidth/latency budgets for high-performance model-parallel inference.
• Collaborate with software teams to co-design fabric topology with model parallelism strategies.
• Interface with platform software teams for fabric software bring-up and integration.

Skills

HPC networking

GPU interconnects

Fabric topology design

Radiation effects understanding

Collaborative engineering

Education

Bachelor of Science in Computer Science or Electrical Engineering

Tools

NCCL

RCCL

InfiniBand

Job description

Blue Origin seeks a hardware-leaning systems engineer to own the foundational hardware decisions for a high-speed interconnect fabric in a large-scale orbital data center. You will oversee from physical layer up to the software stack, ensuring resiliency to radiation and optimal performance for AI workloads.

You will collaborate with software teams to align fabric topology with model parallelism and participate in fabric bring-up and integration efforts.