> Markdown version of [/jobs/ext/1907856-linux-graphics-engineer](https://www.wearedevelopers.com/jobs/ext/1907856-linux-graphics-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Linux Graphics Engineer - **Company:** NVIDIA Ltd. - **Location:** Indianapolis, IN, United States (Remote available) - **Experience:** Expert - **Salary:** $155,000.0 - $175,000.0 - **Contract:** Permanent contract - **Skills:** Application Programming Interfaces (APIs), Amazon Web Services, C++ (Programming Language), Digital Rendering, Software Debugging, Linux, DirectX (Software), Memory Management, Github, Graphics Software, Open Source Technology, Large Language Models, Gaussian, HLSL, Kubernetes, Low Latency, Video Streaming, Unreal Engine, C++14, Lidar, Docker - **Published:** August 3, 2026 - **Apply:** https://jobs.lever.co/paralleldomain/146b6a99-d81d-4210-9cf1-20a408473b69/apply?lever-origin=applied&lever-source%5B%5D=BuiltInNationwide ## About the Role * Experience. 7+ years in real-time graphics or engine programming, with at least one product you took all the way to ship. * Vulkan depth. Production experience with Vulkan - not a side project. You've written and debugged Vulkan render paths, reasoned about memory allocation and descriptor management, chased down synchronization and validation-layer issues, and shipped the result. * Linux graphics in production. You've shipped graphics software on Linux and are comfortable in that environment end to end: GPU drivers, Mesa or vendor stacks, shader toolchains, compositors, containers, and the debugging tools that actually work there. * Modern C++. Highly productive in a large, real C++ codebase with a serious build system and CI. You can navigate an engine you didn't write. * GPU performance work. Demonstrated ability to profile a frame, identify the real bottleneck, and fix it - on both GPU and CPU sides - using tools like RenderDoc, Nsight, or platform equivalents. * Independence. This role starts with a hard problem and not much precedent to lean on. You can scope it, sequence it, and drive it to completion without close supervision. * Communication. You can explain a thorny graphics tradeoff to a non-graphics engineer in five sentences, and give an honest read on schedule when one is asked for. ## Description * Own the Linux and Vulkan migration. Take our current Windows-based renderer from early build to production quality on Linux on Nvidia graphics: correct output across every camera, lidar, and radar configuration we support, at performance that holds up against our benchmarks. You are the person accountable for this landing. * Close the gap between graphics backends. Diagnose and fix the differences between our DirectX and Vulkan paths - memory footprint, shader compilation and pre-caching, driver behavior, synchronization, and the rendering artifacts that show up on one backend and not the other. * Make correctness verifiable. Build the comparison harnesses and regression tests that prove output equivalence across platforms. * Profile and optimize on Linux. Establish the GPU and CPU profiling workflow on the platform, find the bottlenecks, fix them, and codify the patterns so the team isn't rediscovering them in six months. * Contribute to the rendering roadmap. Beyond the migration, take on the broader rendering work: wide-angle and multi-view rendering, reconstruction-based techniques such as gaussian splatting, lighting work in collaboration with our ML team, and moving computation onto available GPU capacity. * Work across the stack. Partner with our build, infrastructure, and simulation engineers. The renderer doesn't ship in isolation - it ships inside a containerized application, through a build system, onto cloud GPU instances, and every one of those seams has a platform dimension. * Use AI tooling actively. LLM-assisted coding and debugging meaningfully accelerate this kind of work when applied well. We expect fluency here and active contribution to the team's practice., * Unreal Engine on Linux. You've shipped a modern Unreal Engine title (5.5 or later) on Linux with the Vulkan RHI. * Unreal Engine internals. Deep familiarity with UE's RHI layer, render graph, shader pipeline, and the parts of the engine most people never touch. * Open source contribution. Meaningful contributions to Vulkan-adjacent open source - Mesa, DXVK/VKD3D, Proton, SPIR-V tooling, validation layers, or the Vulkan ecosystem more broadly. Developers doing this work tend to know where the standard is going before the rest of us do. * Cross-API porting. Prior experience bringing a renderer from DirectX 11/12 to Vulkan, or the reverse. * Adjacent low-latency graphics. Video streaming, cloud gaming, or remote rendering backgrounds - that work lives close to Vulkan drivers and Linux GPU stacks. * Sensor and simulation graphics. Familiarity with lidar, radar, or camera simulation, physically based rendering for machine perception, or neural rendering and Gaussian splatting. Core Tools Vulkan · Unreal Engine 5 · C++ · Linux · HLSL/SPIR-V · RenderDoc · Nsight · Docker · Kubernetes · AWS · GitHub Actions What Success Looks Like In your first six months, you'll have: * Taken the Linux renderer to production quality, with output validated as equivalent across platforms for our supported sensor configurations. * Closed the performance and memory gap between graphics backends. * Left behind a regression and comparison harness the team trusts, so platform parity stays proven rather than assumed. * Started contributing to the broader rendering roadmap and established yourself as a peer to the rest of the graphics team. What Makes a Great Candidate You go deep. You like problems where the answer isn't documented anywhere and you have to read the driver source, the spec, or the disassembly to find out what's really happening. You hold strong opinions about graphics APIs and can defend them, but you treat your tools as instruments, not identity. You'd rather own something end to end than work a ticket queue. And you're drawn to what happens when real-time rendering gets pointed at machine perception instead of human eyes - the constraints are different, and most of the interesting problems are still open. ## Related Videos - [Accelerating Python on GPUs](https://www.wearedevelopers.com/videos/859-accelerating-python-on-gpus) - [Docker network without Docker](https://www.wearedevelopers.com/videos/1418-docker-network-without-docker) - [How to develop an autonomous car end-to-end: Robotic Drive and the mobility revolution](https://www.wearedevelopers.com/videos/22-how-to-develop-an-autonomous-car-end-to-end-robotic-drive-and-the-mobility-revolution) - [Remote Driving on Plant Grounds with State-of-the-Art Cloud Technologies](https://www.wearedevelopers.com/videos/251-remote-driving-on-plant-grounds-with-state-of-the-art-cloud-technologies) - [Coffee with Developers - Stephen Jones - NVIDIA](https://www.wearedevelopers.com/videos/1303-coffee-with-developers-stephen-jones-nvidia) - [Docker exec without Docker](https://www.wearedevelopers.com/videos/1094-docker-exec-without-docker) ## Related Articles - [Dev Digest 120 - Apple and peers](https://www.wearedevelopers.com/magazine/455-dev-digest-120-apple-and-peers) - [Dev Digest 102 - Race conditions](https://www.wearedevelopers.com/magazine/386-dev-digest-102-race-conditions) - [Dev Digest 129 - Now that's what I call private data!](https://www.wearedevelopers.com/magazine/468-dev-digest-129-now-that-s-what-i-call-private-data) - [Dev Digest 121 - AI goes offline](https://www.wearedevelopers.com/magazine/456-dev-digest-121-ai-goes-offline) - [Highest Paying Tech Companies for Developers](https://www.wearedevelopers.com/magazine/220-highest-paying-tech-companies-for-developers) - [Dev Digest 130 - The Diablo is in the details](https://www.wearedevelopers.com/magazine/470-dev-digest-130-the-diablo-is-in-the-details)