Эта вакансия в архиве
Посмотреть похожие вакансии ↓обновлено 2 месяца назад
AI Network Engineer
Описание вакансии
Текст:
TL;DR
AI Network Engineer (InfiniBand/RDMA): Designing, deploying, and operating high-speed InfiniBand and Ethernet fabrics for GPU AI factories across Europe with an accent on large-scale datacenter networking, transport tuning, and network automation. Focus on configuring BGP, OSPF, and VXLAN, optimizing RoCE and InfiniBand for NCCL and UCX workloads, and troubleshooting packet loss, congestion, and training performance regressions.
Location: GPU AI factories across Europe; recurring datacenter and site-based infrastructure work.
Company
operates GPU AI factories across Europe.
What you will do
- Design and deploy InfiniBand NDR 400G/HDR and high-speed Ethernet fabrics for GPU clusters with 1,000+ nodes.
- Configure and operate Arista, Juniper, and Mellanox/NVIDIA networking equipment.
- Manage BGP, OSPF, and VXLAN overlays and tune RoCE and InfiniBand transport for NCCL and UCX workloads.
- Maintain network automation pipelines using Ansible, NetBox, and Nautobot across multiple sites.
- Troubleshoot performance regressions, packet loss, and congestion during live AI training runs.
Requirements
- 4+ years of datacenter networking experience.
- Experience with InfiniBand or 400G/800G Ethernet at scale.
- Deep familiarity with RDMA, RoCE v2, and GPU training cluster communication patterns.
- Strong knowledge of Linux networking internals, including DSCP, ECN, PFC, and adaptive routing.
- Experience with Ansible, Terraform, NetBox, and CI/CD pipelines for network-as-code.
- Ability to interpret tcpdump, perftest, and ib_write_bw diagnostics and correlate results with application performance.
Музыка
Осень
Поддержка