Cluster Architect
Get new jobs like this one by email. Create a free account for job alerts and to save the jobs you like.
Job Summary
Owning the reference designs for GPU clusters, the full-time Cluster Architect will transform customer requirements into comprehensive cluster designs while working remotely and ensuring optimal performance across compute, fabric, storage, and power interfaces.
Key responsibilities
- Develop standard cluster designs for current and next-generation NVIDIA platforms
- Create configurations, BOMs, and rack layouts tailored to customer workloads and site conditions
- Conduct workload analysis to balance design trade-offs related to latency, bandwidth, and GPU density
Required qualifications
- Proven experience designing and delivering GPU clusters of 1,000+ GPUs
- In-depth knowledge of NVIDIA HGX and NVL72 systems, as well as SuperPOD reference architectures
- Familiarity with InfiniBand, Spectrum-X fabrics, and cluster management tools like Slurm or Kubernetes
- Understanding of rack power, liquid cooling, and facility constraints impacting design
- Experience with NCCL, HPL, and MLPerf benchmarking is a plus
Complete Job Description
The complete job description is available to premium members. Premium membership includes:
Full access to 39,548 remote jobs from human-vetted companies, updated daily
Resume Builder - AI-powered tool to craft, enhance, and tailor your resume to a specific job
Twice-monthly live group coaching and the full Remote Career Center
20% member discount on Career Services
Backed by a 30-day money-back guarantee
Company Company Name
Headquarters Headquarters
Founded Founded
Website
The company description is available to premium members. Go Premium for full access to every job and company profile in our remote jobs database, backed by a 30-day money-back guarantee.
