Power Integrity Co-design Engineer

NVIDIA NVIDIA · Semiconductors · Shanghai, China

NVIDIA is seeking a Power Integrity Co-Design Engineer to architect and deliver di/dt mitigation across silicon, package, board, and platform. This role involves defining product-level voltage noise targets, architecting mitigation strategies, and leading codesign tradeoffs. The engineer will also build and lead Sim-to-Si correlation methodology, model next-gen noise features, and lead bug resolution during bringup. A key aspect is applying AI techniques to accelerate power integrity work, requiring strong engineering judgment to balance AI acceleration with rigorous manual validation.

What you'd actually do

  1. Define product-level voltage noise targets, drive them to closure, and sign them off at shipment.
  2. Architect voltage noise mitigation across the full stack - silicon, package, board, platform - and lead the codesign tradeoffs between them.
  3. Co-design noise features with Speed/Power/Reliability, circuit, power-arch, ASIC, and platform teams. You're the connective tissue across the codesign web.
  4. Build and lead the Sim-to-Si correlation methodology for voltage noise.
  5. Model and prototype next-gen noise features - transient sense, droop response, mitigation IP.

Skills

Required

  • BS, MS, or PhD in EE, CE, or related (or equivalent experience) with 4+ years in silicon power integrity, voltage noise, or PDN
  • Deep expertise in at least one of: di/dt analysis and mitigation, voltage droop, PDN design (die + package + board), transient noise, or decap budgeting.
  • Hands-on silicon experience: bring-up, characterization, correlation — comfortable on a bench with scopes, probes, and DAQ, and in front of a simulator.
  • Strong Sim-to-Si correlation instincts
  • Multi-functional collaboration and spec subject area: able to drive a decision through multiple partners, detail it, and own it through sign-offs.

Nice to have

  • Proof of craft: patents, publications, or reusable methodology you built in power integrity, PDN, or di/dt
  • Hands-on experience with groundbreaking GPU, CPU, or AI accelerator silicon at advanced nodes; multi-rail, multi-domain PDN ownership at SoC level (die + package + board co-optimization in production).
  • A track record of applying ML or AI to noise modeling, transient prediction, droop response, or feature optimization — with the validation rigor to know when the model is wrong.

What the JD emphasized

  • Proven use of AI techniques to accelerate power integrity work
  • Clear judgment is required to know where the model is conscientious and where manual measurement is vital
  • A track record of applying ML or AI to noise modeling, transient prediction, droop response, or feature optimization — with the validation rigor to know when the model is wrong