GPU Cluster Architect

🚀 Hiring: Founding GPU Cluster Architect | AI Infrastructure

📍 San Francisco or US remote, with office and cluster-site travel (employees already have to be based in US)

$52M raised. $245M in announced compute contracts. Two next-generation NVIDIA B300 clusters.

I’m partnering with a Series A AI infrastructure startup building GPU supercomputers and a marketplace where customers can resell unused compute capacity.

Their CTO previously led the deployment of 24,000 NVIDIA H100 GPUs across six US data centers. Now, they’re hiring a founding architect to work directly with the CTO and shape the next generation of their infrastructure.

You’ll own:

→ Translating customer workloads into hardware sizing, GPU selection and network topology

→ Designing, deploying and scaling interconnected production GPU clusters

→ Debugging performance and reliability across hardware, Linux, drivers and networking

→ Automating provisioning, monitoring and remediation

→ Supporting production operations, mentoring engineers and making deployments repeatable

I’d love to speak with engineers who have 5+ years of relevant experience and have personally designed and scaled a production HPC or GPU cluster.

Strong server hardware fundamentals are essential. InfiniBand/RoCE, RDMA, Slurm or Kubernetes experience would be valuable.

Deployments above 1,000 GPUs are a plus.

$200K–$400K+ plus equity

Open to remote set up from US or Canada as long as candidates are open to regular travel to SF.