Senior Compute Platform Engineer, LSF - EDA Infrastructure

NVIDIA•Austin, TX

3d•$184,000 - $356,500

About The Position

NVIDIA is consolidating a dual-scheduler estate onto a single IBM Spectrum LSF platform supporting millions of cores across federated EDA compute cells. This deep-specialist role owns scheduler behavior and provides technical expertise when scheduling latency or other farm performance issues arise. The engineer will guide the design of cell topology and federation as the environment grows, work with an IaC engineer to encode scheduler policy, and partner with CAD and methodology teams on demanding workloads, including high-memory jobs, interactive-versus-batch contention, and tape-out crunch periods.

Requirements

BS or MS in Computer Science, Computer Engineering, or equivalent experience.
8+ years of experience in HPC or large-scale batch compute, including 5+ years with IBM Spectrum LSF.
Demonstrated depth in LSF internals, including debugging scheduler behavior beyond documented guidance and explaining a scheduling cycle from submission to dispatch.
Hands-on MultiCluster experience in a production, multi-site environment.
Strong Linux systems fundamentals.
Experience with system programming languages and scripting in Python, Perl, and shell.

Nice To Haves

Experience working on LSF as a developer or in escalation engineering, rather than only as a consumer.
Experience with LSF integration points such as esub, eexec, elim, submit wrappers, RTM, or the LSF APIs.
Background in semiconductor or EDA compute, including workloads with competing license and job constraints.
Experience migrating a production estate from Slurm, PBS, or Grid Engine without a scheduled outage noticed by users.

Responsibilities

Own scheduler behavior across 15–25 federated LSF cells, including mbatchd and mbschd tuning, scheduling cycle analysis, and contention patterns near host-count ceilings.
Diagnose MultiCluster forwarding problems, including remote queue sizing, forwarding policy, and cross-cluster pending behavior.
Set the technical design for cell topology and federation as the compute farm grows, including decisions about cell boundaries.
Work with the IaC engineer to encode scheduler policy into a configuration schema that works with MultiCluster at scale.
Partner with CAD and methodology teams on workloads involving 500GB+ memory jobs, interactive-versus-batch contention, and tape-out crunch bursts.