Senior GPU Infra Engineer — Customer-Facing (Hybrid/Remote)

Rune; Mountain View (CA); Hybrid; USD 175,000 - 260,000; Full time

Job summary

Rune is seeking a hands-on engineer to own customer-facing technical support for our GPU clusters, covering networking, provisioning, and GPU health. You will diagnose issues, resolve them, and avoid hand-offs by maintaining technical judgment.

You will work on RoCEv2, NCCL, and NCCL benchmarking, while managing driver/firmware compatibility across a fleet. This is a hybrid/remote role with a focus on delivering reliable compute at solar sites.

Qualifications

• Hands-on experience with RDMA fabrics, RoCEv2, ECN/DCQCN, PFC, and DCBX configuration.
• Experience debugging multi-node GPU interconnect performance with NCCL and RDMA benchmarking tools.
• Knowledge of GPU interconnect topology and NUMA/PCIe affinity.
• Experience reading GPU health telemetry and managing driver/firmware across a fleet.
• Strong Linux systems administration and fleet automation (Ansible or equivalent).

Responsibilities

• Serve as the primary technical responder for customer-reported networking, GPU-interconnect, and node issues on live clusters.
• Diagnose multi-node NCCL and RDMA performance regressions and distinguish hypotheses from confirmed findings.
• Configure and validate RoCEv2 congestion control and reconcile NIC settings with switch fabric behavior.
• Diagnose GPU hardware faults using DCGM telemetry and Xid codes, differentiating hardware failures from software issues.
• Own driver and firmware compatibility across the fleet and validate updates.
• Provision, re-provision, and decommission GPU nodes with validation testing.
• Triage edge-network and jumpbox access issues affecting provisioning workflows.
• Write incident updates and post-resolution summaries for customer teams.

Skills

RoCEv2

NCCL

NVIDIA DCGM

Linux admin

Ansible

Tools

nvidia-smi

SSH

Job description

Rune is seeking a hands-on engineer to own customer-facing technical support for our GPU clusters, covering networking, provisioning, and GPU health. You will diagnose issues, resolve them, and avoid hand-offs by maintaining technical judgment.

You will work on RoCEv2, NCCL, and NCCL benchmarking, while managing driver/firmware compatibility across a fleet. This is a hybrid/remote role with a focus on delivering reliable compute at solar sites.