Apply on Employer Site

True3D Labs · 3 months ago

Research Engineer - Graphics

New York, NY, US

Full-time

Onsite

Senior Level, Lead/Staff

$150K/yr - $250K/yr

7+ years exp

True3D Labs is seeking a principal level Research Engineer specializing in graphics. The role involves bridging graphics research with production engines, designing high performance graphics systems, and ensuring robust and scalable rendering technology for real products.

Artificial Intelligence (AI)VideoVideo Editing

No H1B

U.S. Citizen Only

Responsibilities

Architect real time rendering pipelines across rasterization, ray tracing, neural rendering, and volumetric or voxel based representations

Design and implement scene representations including triangle meshes, voxel brickmaps, surfels or splats, signed distance fields, and implicit neural fields

Develop high performance shaders and kernels in HLSL, GLSL, WGSL, CUDA, OptiX, Metal Shading Language, and Triton when appropriate

Implement advanced techniques such as mesh shading, work graphs, bindless resources, descriptor heaps, asynchronous compute, sparse residency, transient memory allocators, tiled and clustered lighting, and GPU driven pipelines

Build state of the art reconstruction and quality systems including temporal reprojection, spatiotemporal denoising, super resolution, upsampling, and reservoir resampling methods such as ReSTIR for direct and global illumination

Own profiling and optimization using Nsight, PIX, RenderDoc, GPU counters, flame graphs, and hardware specific tools to reduce divergence, improve occupancy, and hit strict frame budgets

Integrate rendering with content pipelines and engines including Unreal, Unity, Blender, and WebGPU runtimes and deliver production ready SDKs and services

Build capture, playback, and evaluation infrastructure with strong observability, deterministic replays, and golden image tests

Translate research insights into reliable production components that meet latency and throughput targets for interactive and streaming scenarios

Share expertise through design reviews, mentoring, documentation, and reproducible research artifacts

Qualification

Computer GraphicsReal Time RenderingC++20GPU ProgrammingShader DevelopmentPerformance TuningRendering ArchitectureNeural RenderingCommunicationMentoring

Required

PhD in Computer Graphics, Computer Science, or related field, or equivalent research track record with production impact

Seven or more years building real time rendering or graphics systems, including significant time in fast paced or startup settings

Strong publication record in top venues such as SIGGRAPH, TOG, HPG, EGSR, or equivalent impactful artifacts that are widely used in production

Proven experience shipping high performance rendering technology in engines or products with strict frame budgets for desktop, mobile, or XR

Mastery of C++20 and GPU programming with deep understanding of memory hierarchies, synchronization, explicit graphics APIs such as DirectX 12, Vulkan, or Metal, and modern shader toolchains

Demonstrated ability to take ideas from paper to production with measurable wins in image quality and frame time

Strong systems skills including profiling, performance tuning, reliability engineering, and cost awareness across CPU GPU and network boundaries

Excellent communication and the ability to work across research, engine, and product teams

Preferred

Contributions that are widely used in the community such as open source renderers, libraries, datasets, or benchmarks with visible adoption

Experience in neural and differentiable rendering, 3D reconstruction, volumetric video, SLAM, geometric deep learning, or simulation

Experience building and operating large scale rendering or training jobs on Kubernetes, Slurm, or Ray across public cloud environments and modern GPU hardware

Experience with WebGPU and high performance graphics on the web

Experience with compiler or IR work such as SPIR V, DXIL, PTX, graph capture, or custom scheduling and code generation for GPUs

Track record of mentoring teams and setting best practices for rendering quality, performance, testing, and reproducibility

Patents or awards that recognize technical contributions