This vacancy is archived
View similar vacancies ↓updated 1 month ago
Lead Network Engineer (AI)
Job description
Text:
TL;DR
Lead Network Engineer (AI): Designing and operating large-scale AI cluster network fabrics with an accent on InfiniBand, RoCE, and RDMA performance. Focus on optimizing collective communication, congestion control, and fabric reliability for frontier-scale training workloads.
Company
is building high-performance infrastructure for frontier-scale AI training and inference workloads.
What you will do
- Define long-range architecture for AI fabric evolution, including RoCE, NVIDIA InfiniBand, and emerging optical designs.
- Own congestion control strategies, including PFC, ECN, DCQCN, and queue management for high-performance traffic.
- Drive topology design decisions to improve NCCL all-reduce performance and tail latency stability.
- Establish observability standards for fabric telemetry, packet loss, and end-to-end job impact.
- Lead vendor engagement for switches, optics, and NICs while mentoring senior engineering staff.
Requirements
- Demonstrated experience in hyperscale networking, HPC fabrics, or RDMA systems at large scale.
- Deep expertise in ECMP, adaptive routing, queueing theory, and network telemetry pipelines.
- Proven track record designing and operating AI or HPC interconnects.
- Strong working knowledge of distributed training frameworks and collective communication libraries like NCCL and UCX.
- Ability to lead major technology transitions across multiple hardware generations and mixed-vendor environments.