Solutions Architect

Role Overview

Lead GPU Cluster Solution Architect (US remote)

You will own the technical design behind every GPU cluster build—moving from architecture choices to buildable project plans in partnership with deployment, supply chain, and data center operations.

Responsibilities

• Design GPU cluster configurations (compute, storage, networking) aligned to NVIDIA Reference Architecture for each engagement
• Select and design network topology/fabric, including:
  • InfiniBand
  • RoCE
  • High-speed Ethernet
• Architect connectivity (internet/VPN/firewall and dedicated protected optical circuits)
• Own hot/cold sparing strategy to protect contracted SLAs
• Adapt designs to real-world constraints: site power, cooling, and hardware lead times
• Partner with deployment, supply chain, and data center ops to translate designs into buildable project plans

Requirements

• 7+ years in solutions architecture, network engineering, or systems engineering across GPU/HPC/large-scale compute
• Deep knowledge of NVIDIA Reference Architecture (e.g., HGX, NVL72)
• Hands-on experience designing InfiniBand, RoCE, and high-speed Ethernet fabrics
• Experience designing sparing/spares for mission-critical infrastructure
• Experience with high-speed shared storage (e.g., Weka, VAST, DDN)

Nice to Have

• Experience designing for enterprise clients or neoclouds
• Familiarity with the NVIDIA NCP program and certification process
• Experience with capacity/sparing modeling tools