Tech Stack
Tag name is followed by "@" symbol and proficiency level value.
About proficiency levels:
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
AI @ 6
CUDA
Data Pipelines @ 6
GPU
HPC
LLM @ 6
Python @ 6
Robotics
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Details
NVIDIA is the industry leader in high performance computing, gaming and AI. Their GPUs and SOCs deliver performance and efficiency, revolutionizing fields like cell research, robotics, and more. NVIDIA invented CUDA, and is continuing to expand its AI and compute platform.
The Silicon Co-Design Group (SCG) is where architecture, silicon, systems, and manufacturing converge to bring NVIDIA’s products to life. Every CPU, GPU, and Tegra SoC NVIDIA has shipped in the past four years passed through their toolchain on its way to production, and over 200 product SKUs were optimized during the Blackwell generation. NVIDIA is hiring an engineer to lead the rebuild of that toolchain around AI.
This role focuses on the silicon layer of NVIDIA’s productization work. Their tools take a chip from pre-silicon estimates through to the values that ship in firmware and populate customer specs. The role centers on the simulation and configuration engines that feed firmware, manufacturing, and specification systems downstream—using AI to optimize and automate each step. The work includes translating chip behavior into firmware-ready contracts, building agents to understand chip feature interactions, and developing evals to prevent bad products from shipping.
Responsibilities
- Simulate power controller interplay, voltage-frequency operating points, and binning yields. Build systems that push performance and power efficiency.
- Turn understanding of silicon and firmware behavior into context engineering. Break down silicon product optimization workflows into composable skills, hybrid retrieval stages, and orchestration layers.
- Integrate silicon productization tools into a custom agent harness: define tool registries (CLIs and MCPs), webhooks, trace capture, and human-in-the-loop checkpoints.
- Lead eval-driven development for applied AI in production: perform error analysis on real silicon workflows, automate scorers of hardware reasoning, and add CI regression gates that protect product quality.
- Help set the team’s AI direction. Mentor and grow the engineers on the team.
Requirements
- BS or MS in EE/CE/CS (or equivalent experience) and 8+ years in silicon bringup, firmware, or productization engineering.
- Deployed multiple production Python services and data pipelines, including at least one LLM-backed system that SMEs depend on for everyday work.
- Can read silicon characterization outputs (speed, power, voltage noise, or binning) and understand tradeoffs.
- Have a working opinion of new AI tooling within a week of release, demonstrated by running it.
Benefits
- Eligible for equity and benefits.
Applications for this job will be accepted at least until June 29, 2026.