Description
Join NVIDIA's Power Analysis and Optimization team to contribute to ASIC power analysis, modeling, optimization, and flow automation. You will work closely with architecture, design, verification, and physical-design teams to define power targets for architectural changes and new features, identify power-saving opportunities, and resolve power issues.
Responsibilities
- Develop IP-level power models to estimate the impact of architectural changes and new features.
- Analyze power consumption across different workloads and application scenarios.
- Collaborate with architects and design teams to define power targets and evaluate power, performance, and area tradeoffs.
- Use internally developed flows and industry-standard tools to identify power inefficiencies and optimization opportunities.
- Work with RTL and physical-design teams to implement and validate power optimizations.
- Investigate power regressions and determine their architectural, RTL, workload, or physical-design root causes.
- Develop scripts, automation flows, and AI-assisted solutions to improve the efficiency and scalability of power analysis.
- Explore new methodologies for power modeling, analysis, attribution, and optimization.
Requirements
- Expected to graduate in 2027 with a MS or PhD degree in Electrical Engineering, Computer Engineering, Computer Science, or a related field.
- Strong understanding of digital circuit design, computer architecture, or ASIC design methodology.
- Understanding of power-consumption fundamentals and common low-power design techniques.
- Programming or scripting experience in Python, C++, Tcl, or a similar language.
- Good written and verbal communication skills in English.
Nice to Have
- Familiarity with ASIC power-analysis tools such as PrimeTime PX or PowerArtist.
- Experience with power modeling, workload analysis, power attribution, or low-power design.
- Experience analyzing power, performance, and area tradeoffs.
- Experience developing or integrating AI agents, engineering skills, or AI-assisted automation workflows.