Senior GPU Low Power Architect

NVIDIA NVIDIA · Semiconductors · Santa Clara, CA

This role focuses on designing and optimizing low-power GPU and SOC architectures for NVIDIA's products, including those powering AI, Automotive, Graphics, and Mobile. The Senior GPU Low Power Architect will invent and innovate low power architectural/RTL solutions, drive features and roadmaps for efficiency improvements, and contribute to power and performance modeling for GPUs, CPUs, Switches, and platforms. The role requires MSEE/MSCE or PhD with specialization in Low Power Architectures or energy-efficient design techniques, and experience with industry tools like SystemVerilog RTL, UVM, UPF, and Python.

What you'd actually do

  1. Invent and innovate low power architectural/RTL solutions and drive features and roadmaps for improving efficiency of our Desktop, Notebook and server platforms.
  2. You'll work cross functionally with other System/SW Architects, and Product engineers to understand energy efficiency improvements.
  3. Architect, Design and Verify features to help improve battery life and energy efficiency.
  4. Participate in architectural studies, tradeoff analysis, micro-architecture, and design, planning, verification, and validation efforts.
  5. Contribute to the development of techniques to develop highly accurate power and performance models for our GPUs, CPUs, Switches, and platforms.

Skills

Required

  • MSEE/MSCE or equivalent experience
  • 4+ years proven experience or PhD
  • Specialization in Low Power Architectures or energy efficient design techniques
  • SystemVerilog RTL
  • UVM
  • Verdi
  • UPF
  • VCS NLP
  • Python
  • C++
  • Digital and analog circuits
  • Power measurement equipment
  • Board-level power issues analysis

Nice to have

  • HW/SW/System level interactions
  • Power what iffs and tradeoffs
  • Clock tree power reduction and CG strategies
  • RAM power optimization
  • Directed and random functional testing
  • Writing test plans
  • Directed or random diagnostics

What the JD emphasized

  • MSEE/MSCE (or equivalent experience) and 4 + years’ proven experience or PhD, with a specialization in Low Power Architectures or energy efficient design techniques is essential.
  • Design experience with industry tools such as SystemVerilog RTL, UVM, Verdi, UPF, VCS NLP, Python, C++ are essential.