Senior Solutions Architect, AI Infrastructure Networking

About the role

A senior architect owns the network design for a platform that builds and runs GPU clouds and AI factories. The role mixes research with hands-on lab work and customer facing explanation.

What you will do

• Publish network reference architectures covering data centre topologies from a single rack to clusters of thousands of GPUs.
• Define compute fabrics on InfiniBand and RoCEv2 Ethernet, including rail-optimised and fat-tree or Clos designs.
• Specify front-end, storage and management networks and their links to customer data centres and public clouds.
• Design multi-tenant isolation using InfiniBand partitions, VRFs, EVPN-VXLAN and Kubernetes network policy.
• Design Kubernetes networking for AI workloads with Multus, SR-IOV and RDMA plugins, the NVIDIA Network Operator and Dynamic Resource Allocat

What they look for

• Deep networking expertise paired with strong solution architecture skills.
• Experience with large GPU cluster interconnection and tenant isolation approaches.
• Ability to document scale limits and trade-offs in cost, performance, operability and vendor lock-in.
• Comfort collaborating with architects, partner managers and engineers across many countries and time zones.