System Performance Analysis Engineer, Silicon Co-design Group

NVIDIA NVIDIA · Semiconductors · Shanghai, China

This role focuses on system performance analysis for NVIDIA's GPUs and SOCs, involving large-scale benchmarking, designing automation tools, optimizing HW/SW performance and power management, and researching new architectural features. A key aspect is utilizing AI-driven techniques and machine learning algorithms to enhance performance prediction, power management strategies, and automate data collection/analysis.

What you'd actually do

  1. Drive to perform large-scale benchmarks for competitive analysis and comparative analysis.
  2. Design tools to automate product definitions, data collection, test case execution, and results analysis. Provide detailed data analysis of functionality, performance, and latency.
  3. Develop methodologies to optimize HW/SW performance and power management to maximize Perf/Power and user-experience on AC and on DC power. Partner with the different architecture and design teams teams on improvement areas.
  4. Research new architectural features for next generation GPU, CPU and SOC to optimize performance and power features development, and transcend product goals.
  5. Tackle multi-variable problems via system level modeling, testing and characterization, trend analysis/projection, and model verification. Brings depth and expertise in product performance-per-watt feature development.

Skills

Required

  • BS or MS in EE/CE
  • 5+ years of experience in performance and power management
  • Experience with DVFS, power gating, clock policy, sequencing, throttling, control systems, boot/reset
  • Strong EE fundamentals
  • Understanding of digital design, low power design, DVFS, control systems, power management techniques, timing, and architecture
  • Understanding of firmware/driver structures and HW/SW interaction
  • Problem-solving skills
  • Collaboration skills
  • Communication skills

Nice to have

  • Hands-on lab experience (oscilloscopes, multimeters, logic analyzers)
  • Silicon bringup exposure
  • Cross-functional leadership
  • System-level intuition
  • Experience with modeling and simulation tools for performance validation
  • Familiarity with the latest advancements in AI and their applications in power and performance optimization
  • Contributions to open-source projects or communities in low-power design methodologies

What the JD emphasized

  • BS pr MS in EE/CE, 5+ years of experience with a strong focus of performance and power management on one or two of below areas that optimize power for perf-per-watt efficiency in datacenter or client systems: DVFS, power gating, clock policy, sequencing, throttling, control systems, boot/reset.
  • Strong EE fundamentals on digital design, low power design, DVFS, power management techniques, timing, and architecture.
  • Solid understanding of firmware/driver structures and HW/SW interaction with different IPs, including their impact on power, latency, and reliability.
  • Leverage AI to perform advanced optimization accelerate analysis, workflows, generate and review scripts, synthesize technical documentation, and explore solution spaces faster.
  • Utilize machine learning algorithms to predict power consumption and performance based on workload characteristics.
  • Implement AI-driven tools for automating data collection and analysis during validation and characterization phases.
  • Explore creative ways to integrate AI insights into continuous improvement of power and performance metrics.