Senior HPC Cloud Engineer

Overview

Senior HPC Cloud Engineer on NVIDIA's HPC Infrastructure team, owning build-out and daily operation of VLSI high-performance compute environments in the public cloud and serving as primary architecture reviewer and performance lead for new data center clusters across multiple sites.

What you'll do

• Own the build-out and day-to-day operation of VLSI high-performance compute environments in the public cloud.
• Bring up and operate large compute clusters on Google Cloud end to end, including instance provisioning, machine images, OS pipelines, networking, identity and quota management.
• Build automation that makes cluster bring-up reproducible, including infrastructure-as-code, configuration management, image build pipelines, and host lifecycle automation with hostname and identity preservation across maintenance events.
• Extend the same patterns to new on-premises data centers so cloud and on-prem run on one toolchain.
• Work with networking and storage teams on cross-site connectivity, latency budgets, and what is cross-mounted versus kept site-local.
• Debug at depth when clusters misbehave, including image and boot issues, mount failures, scheduler-host interactions, and capacity and quota problems.
• Partner with scheduling and operations engineers so what you build is operable by the whole team.

What you'll need

• B.E./B.Tech/M.Tech in Computer Science, Electrical Engineering, or equivalent experience.
• 8+ years in infrastructure engineering, with substantial time running production compute at scale.
• Deep, hands-on expertise with a major public cloud; Google Cloud preferred, covering compute, networking, IAM, storage, quotas, and operating thousands of instances rather than tens.
• Strong infrastructure-as-code and configuration management skills (Terraform, Ansible, Puppet, or equivalent).
• Experience with machine image build and OS lifecycle pipelines for large fleets.
• Solid Linux systems administration and troubleshooting: filesystems, NFS mounts and latency, networking, boot and image pipelines.
• Working understanding of data center and network fundamentals for compute-intensive workloads, including topology, bandwidth and latency characterisation, and the practical limits of cross-site operation.
• Familiarity with batch schedulers (LSF, Slurm, or equivalent) and how compute nodes join, leave and are managed within them.
• Strong Python automation skills.
• Experience bringing a new environment, cluster or data center online end to end.
• Ability to work effectively across time zones.

Nice to have

• Semiconductor/EDA exposure.

Details

• Location: India, Bengaluru.