Get a reply from this employer
Your resume and cover letter, tailored to exactly what they’re hiring for.
Larsen & Toubro (L&T) in India seeks a senior engineer to deliver and operate ultralow-latency network fabrics for GPU training at scale. You will design east–west InfiniBand and north–south Ethernet connectivity, and optimize leaf-spine topologies with ECMP to ensure deterministic performance.
The role emphasizes reliability, incident management, and automation. You will tune RoCE v2, implement QoS, monitor NCCL performance across pods, and lead P0/P1 response for GPU infrastructure,
Deliver and operate ultralowlatency, highbisection network fabrics for GPU training at ~10K GPU scale with deterministic performance.
(degree, training, or certification required)
BE/B-Tech or equivalent with Computer Science or Electronics & Communication
(no. of years of technical, functional, and/or leadership experience or specific exposure required)
NVIDIA UFM/NetQ (or equivalent), Spectrum/Quantum families, Cumulus/Onyx/SONiC; iperf/NCCLtests; sFlow/telemetry; Ansible/Terraform.
NVIDIA Networking; CCNP/CCIE DC; JNCIPDC / Arista ACE.
Fabric throughput/latency SLOs, congestion events, NCCL job throughput variance, change success rate.
Onsite for build/turnups; oncall for P1 network incidents.