By submitting your resume, you’re expressing interest in our 2027 RDSS (Research and Development Substitute Services) program. Please confirm your eligibility with the local district office before applying the role.

Most engineering careers start with tasks someone else defined, on problems someone else identified. This one doesn't.

NVIDIA has reinvented computing more than once — the GPU, deep learning, the infrastructure that runs modern AI. Today, as accelerated computing reshapes every industry from healthcare to robotics to scientific discovery, the silicon that powers it has never mattered more. The Silicon Co-Design Group is where design intent becomes silicon reality, across consumer, professional, server, and automotive products. As a new graduate here, you will be in the lab on real silicon from day one — owning work, not shadowing it — and when the silicon teaches you something the model didn't predict, your findings will shape the features that go into the next chip.

What you'll be doing:

• Own characterization of pre-production silicon — speed, performance, power, yield, and quality — from early bring-up through production sign-off.

• Define and refine features. Evaluate new silicon capabilities against real-world use cases, quantify the benefit and cost of architected features post-silicon, and provide the data that determines what gets built into future products.

• Build methodologies that last. Develop characterization and validation methodologies that close the simulation-to-silicon gap, correlate measured behavior to design, and scale across products and generations.

• Drive failure analysis. Form hypotheses, design experiments, trace root causes across circuit, firmware, and software layers, and don't stop until the answer is confirmed.

• Analyze real-world workloads — LLM inference and training, generative AI, AAA games, autonomous driving stacks — to push best-in-class performance and energy efficiency.

• Build and deploy AI-driven analysis flows, intelligent data pipelines, and automated triage tools that expand coverage and compress cycle time.

• Build tools the team runs on — automation for characterization, data collection, test execution, and analysis. What you build will be used.

What we need to see:

• BS, MS, or PhD in Electrical Engineering, Computer Engineering, Computer Science, or Systems Engineering.

• Solid fundamentals in digital design, computer architecture, circuit analysis, signal integrity, and statistics — the kind you can apply under pressure, not just in coursework.

• Proficiency in Python, C/C++, or equivalent. You build things. You don't wait for a tool to exist.

• The instinct to own outcomes. When something is wrong, you notice. When you notice, you act.

• You don't need to check every box. We care more about how you think, how you learn, and what you've built than whether your background maps perfectly to this list. If this role excites you, apply.

Ways to stand out from the crowd:

• You've built something with AI, LLMs, or agentic workflows that solved a real engineering problem.

• You've developed a methodology or framework that someone else adopted — your work became the standard, not just the solution.

• You've evaluated a feature, weighed benefit against cost, and produced a recommendation that influenced a real decision.

• Lab experience with oscilloscopes, logic analyzers, BERTs, or similar.

The chips your team works on power the world's most advanced GPUs, workstations, datacenter AI infrastructure, and autonomous vehicles. The engineers who built them started exactly where you are. The difference is they chose to start here — where the ownership is real, the methodologies you build shape future programs, and the features you define ship to the world.

If that's where you want to begin, we want to hear from you.