HPC Architect

The Role:

• You own the technical relationship with compute providers, vetting their clusters and defining the quality bar.
• You work closely with procurement and SREs to ensure only qualified clusters join the network.
• You are the first HPC Architect, building the function from the ground up.

What You’ll Do:

• Vet prospective providers and assess their architecture, hardware, and fabric against Andromeda's metrics.
• Build the acceptance test suite and quality thresholds, formalizing what currently exists as tribal knowledge.
• Guide providers through onboarding and remediation to bring clusters up to standard.

What We’re Looking For:

• Deep HPC experience with GPU clusters at scale, strong fabric knowledge (InfiniBand and RoCE).
• Experience with distributed orchestration (Slurm, Kubernetes) and data-center literacy.
• The ability to write testable standards and deliver critical feedback to providers.