Principal Game Designer

Speed up Unreal Engine NNE inference with NVIDIA TensorRT for RTX runtime. Apr 30, 2026 Neural network techniques are increasingly used in computer graphics to boost image quality, improve performance, and streamline content creation. Approaches like super resolution, denoising, and neural rendering help real-time engines work more efficiently, offering new creative possibilities while keeping performance in mind. Unreal Engine 5 (UE5) has taken several steps in this direction with the introduction of the Neural Network Engine (NNE), which serves as an abstraction layer that unifies inference workloads across multiple backends. Developers can use various runtimes on a GPU or fall back to a CPU depending on available hardware for seamless integration of neural network features in real-time graphics workflows. This blog post covers the new plugin that adds NVIDIA TensorRT for RTX as an NNE runtime option (NNERuntimeTRT) for efficient inferencing on NVIDIA RTX GPUs. To show its benefits, I'll use a simplified UE project that runs a post-process AI model to highlight gains over other GPU runtimes, like DirectML. First, let's briefly discuss the different components involved in the project. TensorRT for RTX overview. TensorRT for RTX enables users to deploy AI models on RTX GPUs more efficiently. It uses a Just-In-Time (JIT) optimizer within the runtime to generate inference engines tailored to the user's GPU. This compilation occurs once on the user's machine and optimizes the model for their specific hardware. As a result, TensorRT for RTX can offer higher throughput compared to default execution providers. For example, throughput comparisons across various models show improvements when using TensorRT for RTX versus DirectML, as measured on an NVIDIA GeForce RTX 5090 GPU. TensorRT for RTX is only compatible with NVIDIA RTX GPUs, from the Turing generation (compute capability 7.5) up to the NVIDIA Blackwell generation (compute capability 10.0). Unreal Engine neural network engine overview. NNE supports multiple runtimes for invoking inferencing tasks and choosing between CPUs and GPUs. TensorRT for RTX is for GPUs and this overview focuses on NNE GPU runtimes. NNE can run inference on the GPU, either: * Synchronously from the CPU, requiring memory synchronization. * Asynchronously through the Render Dependency Graph (RDG), aligning with frame rendering. The synchronous method works well for editors and event-based inference tasks like LLMs, where copying data between host and device is not a concern. In contrast, RDG ties model evaluation to rendering resources, making it ideal for AI post-processing, upscaling, or denoising. The NNE TensorRT for RTX plugin supports both GPU and RDG methods, offering flexibility for various AI applications such as rendering, animation, language, and speech while maintaining strong performance on consumer-grade devices. The style transfer post-processing sample project. I built a basic UE5 project to test the NNE TensorRT for RTX plugin, which applies style transfer models during post-processing. For testing, I set up a simple level using a few basic primitives and a fixed camera to keep the visuals consistent while switching between DirectML and TensorRT, making it easier to compare both results and performance. Prerequisites. While the project is nearly ready for use, having experience with UE5, post-process materials, and engine source compilation is recommended. To run style transfer inference and manage rendering resources, you need an NNE implementation. UE5 offers this through the Neural Post-Processing plugin, which performs inference on the GPU through RDG. For this project, Andrew Ferguson'll utilize the NNERuntimeTRT RDG method. All that's required is to train or download an appropriate style model. Andrew Ferguson'll use a pre-trained one from the ONNX zoo. The project already includes an imported model (candy-9-720.uasset). Project setup. Although the NNE TensorRT for RTX plugin is compatible with the 5.7 binary engine release available from the launcher, the neural post-processing plugin contains a hard-coded list of runtimes within its neural profile asset. It's necessary to update its code to include the NNERuntimeTRT in the list of available runtimes for the neural profile asset. Get started: * Get the engine source from GitHub. If it is your first time accessing the engine code base, you need to link your GitHub account with your Epic account. Review this document, which explains the process for accessing the engine code base: * After initial engine setup, you should have a Visual Studio solution. Open the solution and locate the neuralprofile.h/cpp files under: Engine\Source\Runtime\Engine\Classes\Engine * In neuralprofile.h, add NNERuntimeTRT to ENeuralProfileRuntimeType (line 59) like so: | UNNERuntimeRDGTensorRT UMETA(DisplayName = "NNERuntimeTRT"), | | UNNERuntimeRDGTensorRT UMETA(DisplayName = "NNERuntimeTRT"), UENUM(BlueprintType) enum class ENeuralProfileRuntimeType: uint8 { NNERuntimeORTDml UMETA(DisplayName = "NNERuntimeORTDml"), /** Does not have full operator support*/ NNERuntimeRDGHlsl UMETA(DisplayName = "NNERuntimeRDGHlsl"), UNNERuntimeRDGTensorRT UMETA(DisplayName = "NNERuntimeTRT"), MAX UMETA(Hidden) }; | 5. In neuralprofile.cpp, find the GetNeuralProfileRuntimeName and add to the kRuntimeNames array (line 28): | ,TEXT("NNERuntimeTRT") | 6. The full kRuntimeNames array should be: | static const TCHAR* const kRuntimeNames = { TEXT("NNERuntimeORTDml"), TEXT("NNERuntimeRDGHlsl"), TEXT("NNERuntimeTRT") }; | * Get the plugin from Fab or the NVIDIA Developer page. * Unpack the plugin under the engine plugins folder in this path: ...\Engine\Plugins\Runtime\Nvidia It's also possible to place the plugin in the project, but for its case, it's simpler to keep it under the engine plugins. * Build the engine. Use these detailed instructions for compiling the engine. The first compilation takes time, so consider taking a break or grabbing a coffee. * Get started with the sample project. * Clone the sample project repository. * Load the project with the compiled engine and play the test level (LVL_PPStyleTest). Performance profiling. While playing the test level, activate the engine stats and alternate between TensorRT for RTX (TRT) and DirectML (DML) to assess performance improvements. For more comprehensive profiling, use Unreal Insights to capture frame information and breakdowns for both DML and TRT. On a system with NVIDIA GeForce RTX 5090 GPU at 1080p, in Unreal Insights, DML required 5.7 ms, whereas TRT completed in 3.8 ms - a 1.5x performance improvement. Using other style transfer models. You can use any style transfer models from the ONNX zoo or train your own. Note that ONNX zoo models have fixed dimensions of 1x3x224x224 for both input and output tensors. The neural process plugin enables tiling for small models to larger frame buffers. While this gives visually acceptable results, it's not recommended as it spawns multiple inference tasks per frame, causing frequent context switches between NVIDIA CUDA and graphics. To avoid the extra overhead of multiple context switches, change the model dimensions to 1x3x720x720 to run the inference without tiling while maintaining good visual quality. In the sample project repo, I've included a Python script that resizes the input and output tensor dimensions for style transfer ONNX models. About Homam Bahnassi Homam is a principal DevTech engineer with three decades of experience in games and engine development. He worked on several AAA titles and contributed to Unreal Engine enterprise features. Currently working at NVIDIA on the latest raytracing and AI technologies in Unreal.

Epic set to release Arc Raiders-like extraction shooter with Disney characters by the end of the year. Apr 12, 2026 Apr 11, 2026 Epic Games will allegedly release an Arc Raiders-like shooter later this year. As reported by Bloomberg (thanks, Game File), the Fortnite maker is working with Disney to develop a "shooting game along the lines of Embark Studios' hit Arc Raiders, but with Disney characters battling enemies until they can reach an extraction point". Bloomberg claims that sources close to the development believe the gameplay mechanics lack originality, but "some of the employees are optimistic that Epic will get it right by the launch date". The Walt Disney Company acquired a $1.5bn USD stake in Epic Games back in 2024, launching a multi-year partnership to create an "expansive and open games and entertainment universe connected to Fortnite". Explicit mention was made of Disney, Pixar, Marvel, Star Wars, and Avatar, with users able to "create their own stories and experiences, express their fandom in a distinctly Disney way, and share content with each other in ways that they love". Bloomberg adds that it's thought this Disney deal will result in "at least" two more games, although the second game has had "middling" early internal reviews. The team working on a third game were reportedly reallocated to speed up production, although Epic's senior director of global communications, Liz Markham, disputed Bloomberg's reporting, claiming it is "not reflective of the ambitions of the Disney collaboration. GamesNu is building a new games and entertainment universe of Disney experiences". "Epic's timelines are aggressive and always have been. We've heavily moved developers onto projects with releases approaching, while smaller prototyping teams are working on further-off projects," Markham added. A Disney spokesperson said: "We remain focused on our long-term collaboration with Epic which continues to have strong momentum, and our work to build a transformational games and entertainment universe remains unchanged". On 24th March, Fortnite developer Epic Games announced mass layoffs that affected over 1,000 workers, including key veterans behind the hit online game. A few days later, it was revealed that Mike Prinke, a programmer with seven years of experience a the company, was among the cuts. Sweeney has since confirmed Epic is in contact with the family and will handle the insurance for them.

Fortnite Disney extraction shooter is coming - here's everything Techwiser Media Private Limited know. April 11, 2026 Key highlights. * 1 Epic Games is reportedly building a Disney-themed extraction shooter set to launch in November 2026. * 2 The game will feature Disney characters fighting enemies and racing to reach an extraction point. * 3 Epic and Disney have not officially confirmed the game, but four current and former employees backed the report. Table of contents. A new report says Epic Games is working on a Disney-themed extraction shooter, and it could change how Fortnite looks in 2026. Bloomberg shared the news, based on info from four current and former Epic employees. This one looks like a full game built around Disney characters with a new style of gameplay Fortnite hasn't used before. What is the Fortnite Disney extraction shooter? According to a Bloomberg report, Epic Games is working on a game featuring Disney characters in an extraction shooter format, similar to Arc Raiders. You drop into the game, fight through enemies, and try to make it to an extraction point before other players eliminate you. Which Disney characters you'll play as hasn't been confirmed yet. But Epic already has deals with Marvel, Star Wars, and Pixar, so the roster could end up being massive. When is it coming out? The game is targeting a November 2026 release. That's a tight window, and some people inside Epic are actually worried. Internal reviewers have flagged that the game mechanics feel"not very original" right now, though other team members are optimistic that it'll come together. It's worth remembering that Fortnite's Battle Royale wasn't an original concept either, and it actually turned out fine. Is it part of Fortnite or a separate game? This is one of the biggest unanswered questions right now. Nobody has confirmed whether this will be a standalone game or something you access from inside Fortnite. Epic has done both before. Fortnite already has modes like Rocket Racing and LEGO Fortnite that feel like completely different games, but launch from the same menu. This Disney extraction shooter may follow that same model, but it could also be its own separate release entirely. Either way, it's tied to the $1.5 billion investment Disney made into Epic back in 2024 because that deal was always supposed to produce more than just a character skins collaboration. Unfortunately, Techwiser Media Private Limited all know that Fortnite's growth has slowed down, and Epic has been very open about that. CEO Tim Sweeney said the company has struggled to keep delivering what he called "Fortnite magic" every season, and the 1,000 employee layoffs were partly due to that. This Disney extraction shooter could be a big bet on getting that momentum back. If Epic can pull off a fun, polished game with recognizable Disney characters in a genre that's been growing fast, it could bring in a completely new audience. Keep an eye on any official announcements from Epic or Disney in the coming months. November isn't that far away.