NVIDIA is building some of the most sophisticated semiconductor platforms globally, powering breakthroughs in networking, AI infrastructure, and high-performance computing. Our NICs and Switches link the largest AI clusters worldwide. Powered by the AI revolution we are changing how hardware architecture work happens!
We strive to define and implement this new agentic hardware architecture, and we are seeking an experienced architecture lead for our AI-for-Architecture team within the Networking Architecture group. You will define and promote intelligent architecture flows across our NIC and Switch architecture groups. You will develop the agents, tools, and processes that transform how architects explore design tradeoffs, and validate models and specifications. In this leadership role with org-wide scope, you will collaborate with an AI engineering team to develop flows from concept to production. You will also serve as the technical authority on what good AI architecture work means at NVIDIA.
What You'll Be Doing:
- Define the roadmap for AI-driven architecture flows — from data collection, modelling and micro-architecture agents that help architects explore feature options, to review and validation agents that set the standard for output quality.
- Work with architects on tough problems, suggest and build agentic flows to address these problems and shorten the time to a good solution.
- Partner with the AI engineering pod to translate architecture workflows into production agents, MCP integrations, and eval harnesses.
- Act as the domain authority and quality judge: recognize what excellent architecture output looks like and verify that AI-assisted flows meet that bar.
- Drive adoption: work with networking architects, run beta cycles, close feedback loops and facilitate widespread usage of our tools/methods.
- Represent the team's technical direction to senior architecture and product leadership.
What We Need to See:
- B.A, M.Sc. or Ph.D. in Computer Engineering, Electrical Engineering, Computer Science, or equivalent experience.
- 6+ years in hardware, firmware, or system architecture — NIC, switch, DPU, CPU, or SoC. Experience with architectural workflows defining microarchitecture specs, performance models, or architecture decision documents, etc..
- Ability to understand and adopt agentic AI workflows and tools (Claude/Codex/Cursor).
- Engineering committed to quality — for example, the ability to judge AI output quality for architecture tasks and suggest validation-fix strategies to improve it.
- Strong interpersonal skills — explain AI flows for architects lacking AI engineering expertise, and distill hardware architecture trade-offs for engineers without architectural backgrounds.
- Consistent track record driving adoption of new tools or processes across a technical organization.
Ways to Stand Out from the Crowd:
- Hands-on experience applying LLMs, agents, or prompt-plus-code pipelines to real engineering work.
- Expertise with high-speed networking silicon — InfiniBand, Ethernet, switch fabric architecture, NIC/RDMA subsystems.
- Familiarity with LLMs: transformer architecture (attention, MLP, MoE) and LLM training/inference parallelism (DP, PP, EP, TP).
- Background defining and measuring engineering productivity metrics — making the impact or our work visible to leadership.
- Track record shipping internal platform tools or developer-experience infrastructure at scale.
NVIDIA's architecture teams are among the most technically demanding in the industry. If you're a senior architect who has already started using AI — and wants to institutionalize that change across an entire organization — we want to hear from you.
