Principal AI Network Architect
Core
Architect next-generation scale-up switches and define network architectures for AI clusters, racks, and multi-rack domains to support AI training and inference workloads.
Role type
Principal AI Network Architect
Builds
End-to-end AI infrastructure solutions including switch silicon, optics, interconnect technologies, and control planes.
Domain
Hardware engineering, AI infrastructure, high-performance networking
Deliverable
production ML models | product features | infrastructure
Required skills
System-level architecture for networking, compute, storage, and accelerators; AI scale-up and scale-out networking design; High-performance switch architecture; ASIC and SerDes technology expertise; Optical networking technologies (OCI, XPO, CPO, OBO); Architecture tradeoff analysis; Distributed systems and collective communication patterns; Rack integration and thermal management.
Preferred skills
Experience with RDMA/RoCE, SONiC, and network operating systems; Knowledge of AI workload patterns and inference systems.
Technologies
Ethernet, RDMA/RoCE, SONiC, switch ASICs, SerDes, retimers, PHYs, OCI, XPO, CPO, OBO
Responsibilities
Define network architectures for AI clusters and multi-rack scale-up domains; Lead architecture studies for switch silicon, optics, and control planes; Develop performance, capacity, reliability, and TCO models; Provide technical direction to switch development teams and partners; Evaluate emerging networking technologies and influence industry roadmaps.
Seniority
Principal, hands-on IC with strategic leadership
