Principal Network Engineer for high-speed network fabrics in AI datacenters.
•Provide primary support and maintenance for high-speed network fabrics in Graphcore AI datacenters, collaborating across architecture, hardware, software and infrastructure teams.
•Key Responsibilities Maintain and optimize network fabrics and topologies in Graphcore datacenters.
•Collaborate with network architects from design through deployment.
•Provide primary escalation for hardware, link, and performance issues.
•Triage day-to-day topology and fabric issues for scale-up/scale-out GPU clusters.
•Requirements 9+ years network engineering experience and 3+ years in AI or hyperscale datacenters.
•Knowledge of RDMA (ROCEv2/InfiniBand) and lossless Ethernet (ECN, DCQCN).
•Experience with high-bandwidth fabrics (400G/800G) and multi-tier CLOS networks.
•Experience troubleshooting routing (BGP, OSPF) and VXLAN implementations.