HPC / GPU Systems Engineer

Role overview Investigate difficult GPU and high-performance computing incidents, turning incomplete or conflicting signals into evidence-backed assessments and practical next steps. The work spans GPU nodes, high-speed networks, storage, Linux, and workload schedulers, with a focus on restoring productive compute time while managing operational risk.

Responsibilities - Diagnose GPU node faults across drivers, firmware, hardware, and runtime signals, using raw logs and tools such as NVIDIA system management and DCGM diagnostics. - Isolate InfiniBand or RoCE network faults down to links, ports, transceivers, or cabling, and prepare evidence for service providers or vendors. - Trace storage, Linux, and scheduler issues affecting multi-GPU workloads. - Critically review diagnostic results, correct false leads, and improve incident rules and runbooks. - Plan live-system changes with risk assessments, validation, and rollback steps. - Participate in escalations and an on-call rotation when its scope and schedule are agreed.

Requirements - Hands-on experience investigating GPU, HPC, or large-scale Linux infrastructure incidents. - Working knowledge of NVIDIA drivers and runtime signals. - Familiarity with RDMA networking such as InfiniBand or RoCE. - Experience operating Slurm or Kubernetes workloads and interpreting telemetry critically. - Clear communication with engineers, customers, and infrastructure partners, plus sound judgment when changing business-critical systems.

Nice to have - Python or Bash for operational tooling. - Infrastructure automation experience. - Ceph or parallel file systems experience. - Data center operations experience.

Benefits and work setup - Full-time role with remote work accepted across North America, South America, Europe, Africa, South Asia, and East Asia; US hubs are preferred.