Energy Modeling Engineer: Power Analytics for AI GPUs

NVIDIA Corporation; Austin (TX); On-site; USD 116,000 - 219,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA Corporation in Austin, TX seeks an Architecture Energy Modeling team member to study and implement energy modeling for GPUs, CPUs, and Tegra SOCs. You will collaborate across Architects, ASIC Design, Low Power, Performance, Software, and Physical Design to improve power management.

You will develop data movement power models and visualization tools (PowerBI/OpenSearch), applying ML to answer design questions and set energy targets for next-gen chips.

Qualifications

• Pursuing or completed MS/PhD in Electrical or Computer Engineering or equivalent.
• Strong coding skills in Python and C++.
• Ability to analyze algorithms and discuss runtimes and memory.
• Understanding of VLSI, digital design, and computer architecture.
• Background in power/energy concepts and low-power design is a plus.

Responsibilities

• Define and implement tools and methodologies for energy modeling from post-layout data.
• Develop tools to sanitize model metrics for high correlation accuracy.
• Integrate power models with performance tools and data movement analytics.
• Identify bottlenecks in flows to speed up model delivery.
• Mine data from pre- and post-silicon runs to find key data paths.

Skills

Python

C++

Algorithms

VLSI concepts

Computer architecture

Machine learning

Power modeling

Data analysis

Education

MS or PhD in Electrical or Computer Engineering

Tools

PowerBI

OpenSearch

Job description

NVIDIA Corporation in Austin, TX seeks an Architecture Energy Modeling team member to study and implement energy modeling for GPUs, CPUs, and Tegra SOCs. You will collaborate across Architects, ASIC Design, Low Power, Performance, Software, and Physical Design to improve power management.

You will develop data movement power models and visualization tools (PowerBI/OpenSearch), applying ML to answer design questions and set energy targets for next-gen chips.