New opportunity

Senior Physical Design AI Engineer

NVIDIA · Bengaluru, India

About this role

NVIDIA is seeking a Senior Physical Design AI Engineer to join their Physical Design AI team in Bengaluru. This role focuses on combining deep physical design expertise with modern AI technologies to develop intelligent, scalable solutions for physical design implementation. You will collaborate with domain experts to address complex implementation challenges and build agentic AI solutions that enhance productivity, efficiency, and quality across NVIDIA product lines. Responsibilities include delivering token-optimized AI solutions, maintaining and enhancing existing AI tools, benchmarking AI models, researching new AI methods for physical design, and building regression testbenches to validate AI solution performance. Required qualifications include 4+ years of experience in physical design, a BE/BTech/MTech degree or equivalent, and a strong understanding of the RTL-to-GDSII flow, including synthesis, place and route, clock tree synthesis, timing convergence, and layout closure. Candidates must have a solid grasp of AI concepts such as tokens, KV cache, context windows, agents, subagents, and hooks, along with hands-on experience using AI tools like Claude, Codex, and Cursor. Proficiency in Tcl, Python, and industry-standard place-and-route tools is required. Experience with Synopsys ICC2 and Cadence Innovus is preferred. The role requires strong analytical, problem-solving, and communication skills, with the ability to work effectively in a global, hybrid environment.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Physical DesignRTL-to-GDSIISynthesisPlace and RouteClock Tree SynthesisTiming ConvergenceLayout ClosureAgentic AIClaudeCodexCursorSynopsys ICC2Cadence InnovusTclPython

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Developing Agentic AI for Physical Design
The core mandate is to bridge deep physical design expertise with AI to build intelligent, scalable solutions. You must translate complex RTL-to-GDSII implementation challenges into agentic workflows that improve productivity and quality across NVIDIA product lines.
Optimizing AI Model Performance
You are responsible for delivering token-optimized AI solutions. This requires balancing model effectiveness with efficiency, including benchmarking performance across different AI models and managing technical constraints like context windows and KV cache.
Validating AI Solutions via Regression
Building robust regression testbenches is essential to ensure that AI-driven design outputs meet strict quality and reliability standards. You must demonstrate how your AI tools maintain consistency in timing convergence and layout closure.
Automation and Tool Integration
You must leverage Tcl and Python to integrate AI solutions into standard industry flows, specifically Synopsys ICC2 or Cadence Innovus. Proficiency in scripting is required to automate complex physical design tasks effectively.

Likely challenges to think through

Bridging Domain Expertise and AI Implementation
A significant challenge is translating nuanced physical design problems into structured AI prompts or agentic logic. Prepare by articulating how you would decompose a complex task like clock tree synthesis into smaller, AI-manageable subagents or hooks.
Maintaining Design Quality in AI-Driven Flows
There is a risk that AI-generated outputs may introduce subtle timing or layout violations. Prepare by discussing your methodology for rigorous validation, including how you define 'success' metrics for AI models compared to traditional deterministic EDA tool results.
Managing Global Hybrid Collaboration
The role requires working with global implementation teams. You will need to balance local development of AI tools with the needs of diverse, distributed teams. Prepare by highlighting your experience in communicating technical requirements and gathering feedback from end-users in a hybrid work environment.

Practice your answers

Technical practice

RTL-to-GDSII Flow Optimization

What to coverExplain how you would identify and resolve a timing convergence bottleneck in a high-performance design. Discuss your experience with Synopsys ICC2 or Cadence Innovus and how you have historically balanced power, performance, and area (PPA) trade-offs.

Agentic AI in Physical Design

What to coverDescribe how you would design an agentic AI workflow to automate a specific stage of the physical design process, such as clock tree synthesis. Address how you would define the agent's 'hooks' and decision-making logic to ensure the output meets strict design constraints.

Token Optimization for AI Models

What to coverWhen integrating LLMs into physical design automation, how do you manage token usage and context window limitations? Explain your strategy for optimizing prompts or data structures to ensure the AI remains efficient and accurate during complex design tasks.

Regression Testbench Development

What to coverHow do you build a robust regression testbench to validate that an AI-generated physical design solution maintains quality and reliability? Discuss the metrics you would track to ensure the AI model's performance does not degrade over time.

Automation with Python and Tcl

What to coverProvide an example of a complex automation task you solved using Python or Tcl within a place-and-route environment. Focus on how your script improved productivity or reduced manual intervention in the design flow.

Benchmarking AI Models

What to coverHow do you approach benchmarking different AI models for a specific physical design challenge? Explain your methodology for comparing model effectiveness, including how you handle data consistency and evaluation criteria.

Managing KV Cache and Context

What to coverIn the context of building AI agents for physical design, how does the management of KV cache impact the performance of your agentic solutions? Suggest how you would architect the agent to handle long-running design sessions effectively.

Behavioural practice

Cross-Functional Collaboration

What to coverDescribe a situation where you had to work with domain experts who were skeptical of an automated or AI-driven approach. How did you communicate the value of your solution and incorporate their feedback to improve the final design outcome?

Handling Complex Technical Challenges

What to coverTell me about a time you encountered a significant roadblock in a physical design project. What was the situation, what specific actions did you take to troubleshoot the issue, and what was the final result?

Adapting to New Technologies

What to coverGive an example of how you have integrated a new AI tool or methodology into your existing workflow. Explain the learning curve you faced and how you measured the success of this integration.

Working in a Global Hybrid Environment

What to coverDescribe your experience working in a global, hybrid team. How do you ensure effective communication and maintain project momentum when collaborating with colleagues across different time zones and working styles?

Sharpen your knowledge

Topics to prepare

Agentic AI Architecture
Understand the lifecycle of AI agents, specifically focusing on subagents, hooks, and context window management. Exercise: Map out a hypothetical agentic workflow that automates a specific stage of the RTL-to-GDSII flow, defining how subagents would handle sub-tasks and how hooks would interact with EDA tool APIs.
Token-Optimized AI Solutions
Study techniques for optimizing LLM interactions, including prompt engineering for physical design constraints and managing KV cache efficiency. Exercise: Analyze a complex physical design script and draft a strategy to convert its logic into a token-efficient prompt or function call that minimizes latency while maintaining output accuracy.
RTL-to-GDSII Implementation Flow
Review advanced concepts in synthesis, place and route, clock tree synthesis, and timing closure. Exercise: Create a checklist of common convergence bottlenecks in modern process nodes and describe how an AI-driven approach could proactively identify or resolve these issues compared to traditional manual tuning.

Skills to revise

AI-Assisted Development with Cursor and Claude
Gain proficiency in using AI-integrated development environments to accelerate code generation and debugging. Exercise: Use Cursor to refactor a legacy Tcl or Python script used in your current physical design workflow, focusing on improving modularity and adding automated error handling.
Regression Testbench Development
Develop skills in building robust validation frameworks for AI-generated outputs. Exercise: Design a regression test suite that compares the PPA (Power, Performance, Area) results of an AI-optimized design against a baseline manual implementation to ensure quality and reliability.
EDA Tool Automation (Synopsys ICC2/Cadence Innovus)
Master the scripting interfaces for industry-standard tools. Exercise: Write a Python or Tcl script that extracts specific timing reports or congestion metrics from an ICC2 or Innovus log file and formats the data for ingestion by an AI model.

Your preparation checklist

Map Physical Design Expertise to AI Workflows
Prepare to discuss how your experience in RTL-to-GDSII, synthesis, and timing closure can be augmented by AI. Identify specific bottlenecks in your past projects where an agentic AI solution could have improved productivity or layout quality.
Demonstrate Proficiency with AI Tools and Concepts
Be ready to explain your hands-on experience with Claude, Codex, and Cursor. Review technical concepts including token optimization, KV cache management, context window limitations, and the architecture of agents, subagents, and hooks.
Prepare a Technical Case Study on Automation
Draft a walkthrough of a complex automation task you solved using Tcl or Python within Synopsys ICC2 or Cadence Innovus. Focus on how you validated your scripts and ensured reliability through regression testing.
Develop Questions on AI Integration Strategy
Prepare thoughtful questions for the interviewers regarding their current AI benchmarking methodology, how they balance token efficiency with design quality, and how they integrate AI-driven insights into existing physical design implementation flows.

Availability last checked: 18 Oct 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…