Staff System Software Engineer, Rtl-to-gds Flow Platform

NVIDIA NVIDIA · Semiconductors · Santa Clara, CA

NVIDIA is seeking a Staff System Software Engineer to build and modernize production RTL-to-GDS flow infrastructure for chip design. This role involves evolving legacy systems into a cleaner, more structured platform, focusing on synthesis, physical design, timing, signoff, and ECO workflows. The engineer will extend configuration systems, improve launch infrastructure, develop pre-launch checks, and build job-control and observability capabilities for complex workflows.

What you'd actually do

  1. Build and modernize production RTL-to-GDS flow infrastructure across synthesis, place-and-route, timing, signoff, ECO, and handoff workflows
  2. Extend YAML/configuration systems to model workflow intent, stage contracts, validation markers, generated artifacts, override order, and backward-compatible project behavior
  3. Improve Make, Perl, Tcl, Python, and related launch infrastructure for generated runsets, EDA tool setup, distributed execution, status tracking, and failure diagnosis
  4. Build faster prelaunch and in-run checks for missing inputs, stale generated files, invalid hooks, broken environment setup, bad constraints, and inconsistent design state
  5. Develop job-control and observability capabilities for hierarchical and internally launched workflows, including parent-child job attachment, logs, provenance, and structured status

Skills

Required

  • B.S. or M.S. in CS, EE, CE, or equivalent experience
  • 12+ years building, modernizing, or operating production EDA, VLSI CAD, RTL-to-GDS, physical design, or large engineering workflow systems
  • Strong hands-on experience with RTL-to-GDS or implementation flows, including setup, generated collateral, tool launch, checks, timing/signoff handoff, and debug workflows
  • Strong Tcl and Make experience in real EDA automation environments
  • Practical software engineering skill in Python, Perl, Go, or C++
  • Ability to reason about layered configuration, includes, overrides, variable expansion, generated outputs, validation state, and compatibility with older projects
  • Excellent Linux debugging fundamentals
  • Track record of improving legacy production systems without breaking active users

Nice to have

  • Background with production flows using commercial synthesis, place-and-route, timing, extraction, DRC/LVS, power, or signoff tools
  • Experience designing workflow engines, runset generators, configuration systems, template-driven automation, or job-control infrastructure
  • Background with distributed schedulers and shared compute environments such as LSF, Slurm, Grid Engine, or similar systems
  • Experience with NFS-heavy shared filesystems, stale state, partial writes, generated input files, provenance tracking, and reproducibility issues
  • Experience building structured checks, validation markers, data-fidelity tracking, dependency tracing, observability, or better-tested legacy Tcl/Make/Perl replacements

What the JD emphasized

  • 12+ years building, modernizing, or operating production EDA, VLSI CAD, RTL-to-GDS, physical design, or large engineering workflow systems
  • Strong hands-on experience with RTL-to-GDS or implementation flows, including setup, generated collateral, tool launch, checks, timing/signoff handoff, and debug workflows
  • Strong Tcl and Make experience in real EDA automation environments, plus practical software engineering skill in Python, Perl, Go, or C++
  • Excellent Linux debugging fundamentals and a track record of improving legacy production systems without breaking active users