Senior Manager, Performance Engineering - Kernel and Software Platforms

Experteer Overview In this role, you guide an engineering team to maintain peak product performance across the full hardware lifecycle. You bridge design-time predictions with real-world execution, spanning pre-silicon to production hardware and post-release support. You shape performance expectations, curate workloads, and align with CUDA release timelines while advancing automation with AI tooling. This is a hands-on leadership position shaping how NVIDIA delivers high-performance DL architectures and AI computing platforms. Compensation / Benefits * Oversee end-to-end performance tracking from pre-silicon through production hardware and maintenance * Build and refine theoretical and empirical performance models and correlate predictions with real hardware telemetry * Evaluate DSL-to-IR performance flows from high-level DSLs (e.g., Triton, PyTorch) to target code across platform maturity * Create and maintain stress-test workloads and testlists to detect regressions and validate releases * Coordinate CUDA release cadence with compiler, architecture, and platform software teams * Drive workflow automation with AI tooling (Claude, Codex, agentic LLM workflows) to streamline telemetry analysis and reporting pipelines Tasks * 10+ years in systems/software performance engineering, benchmarking, or related fields * 5+ years leading or managing technical engineering teams * Experience tracking performance through full hardware lifecycle from design-time to production hardware * Proven ability to build and correlate performance models with hardware telemetry without daily low-level kernel development * Understanding of kernel compilation pipelines and compiler flows (DSL -> IR -> target code) and their impact on execution efficiency * Experience developing workload testlists and aligning performance delivery with major software release cycles (CUDA cadence) * Proficiency in Python and practical use of generative AI APIs/models (Codex, Claude) for automation and analytics Key requirements * equity * benefits * hybrid work arrangement * career growth * competitive salary * relocation assistance

aay_ * Coordinate CUDA release cadence with compiler, architecture, and platform software teams * Drive workflow automation with AI tooling (Claude, Codex, agentic LLM workflows) to streamline telemetry analysis and reporting pipelines Tasks * 10+ years in systems/software performance engineering, benchmarking, or related fields * 5+ years leading or managing technical engineering teams * Experience tracking performance through full hardware lifecycle from design-time to production hardware * Proven ability to build and correlate performance models with hardware telemetry without daily low-level kernel development * Understanding of kernel compilation pipelines and compiler flows (DSL -> IR -> target code) and their impact on execution efficiency * Experience developing workload testlists and aligning performance delivery with major software release cycles (CUDA cadence) * Proficiency in Python and practical use of generative AI APIs/models (Codex, Claude) for automation and analytics Key a

Engineering * equity * benefits * hybrid work arrangement * career growth * competitive salary * relocation assistance