Staff Graphics Engineer, Rendering (Simulation)

General Motors•Sunnyvale, CA

1d•$160,200 - $246,300•Remote

About The Position

As a Staff Graphics Engineer on the Rendering team, you will push the visual and physical fidelity of a GPU-accelerated, physically-based sensor simulator used to generate synthetic data — cameras, LiDAR, radar, and depth sensors — for autonomous vehicle development and validation. This is a deep graphics role: you will own and advance the light transport, materials, atmosphere, and sensor-imaging pipelines that determine how realistic our rendered output is. Your work directly affects the quality of the synthetic data that AV perception and planning teams train and validate against, where physical accuracy is not a "nice to have" but the entire point. We render with NVIDIA OptiX and CUDA, evaluate physically-based materials through NVIDIA MDL, and produce spectrally-derived, radiometrically-grounded imagery. We're looking for someone with a genuine graphics background — the kind of engineer who has implemented a BRDF from a paper, debugged energy loss in a microfacet model, and cares about the difference between physically plausible and physically correct.

Requirements

BS, MS, or PhD in Computer Science, Computer Graphics, or equivalent experience.
8+ years of professional software engineering experience, with a substantial portion focused on real-time or physically-based rendering.
Deep, hands-on graphics expertise: physically-based rendering, BRDF / microfacet theory, Monte Carlo path tracing, importance sampling, and light transport. You can read a SIGGRAPH paper and turn it into correct, performant shader code.
Strong C++ skills (modern C++17/20), including performance optimization, memory management, and clean API and system design in a large codebase.
Significant GPU programming experience — CUDA and/or a ray tracing API (OptiX, DXR, Vulkan RT) — with a real understanding of GPU architecture and execution.
Solid grounding in the math of rendering: linear algebra, sampling and probability, radiometry, and color science.
Experience profiling and optimizing GPU rendering workloads (e.g., Nsight) on real, complex scenes.
Strong communication and collaboration skills, and a track record of driving technical design across teams.
Movie industry experience background

Nice To Haves

Experience implementing or extending a production renderer, game engine rendering core, offline/path-tracing renderer, or commercial graphics product.
Experience with NVIDIA MDL, OSL, or other physically-based material systems.
Experience with OptiX specifically, or other GPU ray tracing / BVH-based pipelines.
Spectral rendering, atmospheric scattering models (Prague, Hosek-Wilkie, Bruneton), or volumetric cloud rendering (Schneider/Nubis-style ray marching, Hillaire multi-scattering).
Camera and sensor imaging: ISP pipelines, sensor noise modeling, lens distortion, motion blur, rolling shutter, exposure and white balance.
Experience with OpenUSD scene description and asset pipelines.
Familiarity with ROS and integrating rendering output into robotics / AV data stacks.
Experience enabling rendering for machine learning workloads — synthetic data generation, reinforcement learning environments, or closed-loop simulation.
Contributions to open-source rendering, graphics, or simulation projects, or relevant publications.

Responsibilities

Advance the light transport in our OptiX-based renderer — improving sampling, multi-bounce indirect illumination, and overall path tracing fidelity while keeping it tractable for large-scale, time-stepped simulation.
Own and extend physically-based materials through our NVIDIA MDL integration — microfacet BRDFs, layered and coat materials, and correct energy conservation in OptiX closest-hit evaluation.
Develop our spectral and radiometric pipeline — wavelength-dependent sky and material response, spectral-to-XYZ-to-RGB conversion, and physically-grounded units from scene radiance through to sensor output.
Push the atmosphere and volumetrics — the spectral sky model (Prague Sky Model), volumetric clouds (multi-layer, multi-scattering, phase functions, aerial perspective), participating media, and fog.
Improve camera image formation — the GPU ISP (demosaic, tone mapping, exposure/EV100, auto white balance via correlated color temperature), and physically-based lens and sensor effects such as motion blur, rolling shutter, lens distortion, and sensor noise.
Design and tune sampling and denoising strategies — importance sampling, blue-noise and stratified sampling, MIS, and spatial/temporal denoising that reduce variance without compromising physical accuracy.
Profile and optimize the per-frame GPU pipeline — OptiX launches, BVH (GAS/IAS) construction, material uploads, and denoising — to render more geometry and richer materials within budget.
Contribute to the OpenUSD scene and material pipeline and collaborate with adjacent teams to plumb new lighting, weather, and sensor parameters end-to-end.
Uphold high standards through technical design documents, code review, reference-image regression testing, and mentorship of other engineers.