Unreal Engine 5.4 brings animation, rendering, and AI upgrades
Epic Games has released Unreal Engine 5.4, packed with new features and improvements aimed at benefiting game developers and creators across industries. The latest version delivers toolsets used internally by Epic to build high-profile releases like Fortnite Chapter 5, Rocket Racing, Fortnite Festival, and LEGO Fortnite.
“Unreal Engine 5.4 is here, and it’s packed with new features and improvements to performance, visual fidelity, and productivity that will benefit game developers and creators across industries,” stated the official Epic Games announcement.
Animation overhaul
One of the major focus areas in 5.4 is animation, with Epic describing “substantial updates” to Unreal’s built-in toolset. This allows developers to quickly rig characters and author animations directly within the engine without external software.
“With an experimental new Modular Control Rig feature, you can build animation rigs from understandable modular parts instead of complex granular graphs, while Automatic Retargeting makes it easier to get great results when reusing bipedal character animations,” Epic explained.
The company says its animation authoring tools have been made “more intuitive and robust” with streamlined workflows. This includes experimental new gizmos, reorganised animator controls, constraint system upgrades, and a layered control rig feature to simplify adding animations atop clips.
Rendering advancements
Nanite, Unreal’s micropolygon geometry system, gains an experimental tessellation feature to add fine details at render time. Software variable rate shading via Nanite compute materials delivers substantial performance gains.
Epic has enhanced its temporal super resolution (TSR) tech with improved stability, reduced ghosting, and new visualisation modes to ease fine-tuning.
The company has refactored rendering systems to improve parallelisation and 60Hz performance. GPU instance culling boosts hardware ray tracing which also gains new primitive types and an optimised path tracer.
AI and machine learning
Unreal Engine’s Neural Network Engine (NNE) moves from experimental to beta status with runtime and in-editor support. NNE allows loading and efficiently running pre-trained neural network models for use cases like tooling, animation, rendering, and physics.
“NNE addresses these disparate needs by providing a common API, enabling easy swapping of backends as required. We’ve also provided extensibility hooks to enable third-party developers to implement the NNE interface in a plugin,” said Epic.
Productivity boosts
Several major features aim to improve developer productivity and iteration speed in 5.4:
- Cloud Derived Data Cache for sharing cached engine data across distributed teams
- Faster local caching using a new Unreal Zen Storage server architecture
- Multi-Process Cook (now production-ready) to leverage extra CPU/RAM for content cooking
- Unreal Build Accelerator (beta) to accelerate C++ compilation across distributed nodes
Media and entertainment updates
Virtual production capabilities see the Virtual Camera tool promoted to production-ready status with Android support added. VR scouting gains an experimental OpenXR toolkit.
An experimental Motion Design mode introduces specialised tools for 2D motion graphics authoring based on feedback from broadcasters.
The USD Importer allows importing garments and simulation parameters from apps like Marvelous Designer to simulate clothing within Unreal in minutes.
You can find the full Unreal Engine 5.4 release notes here.
(Image Credit: Epic Games)
See also: Stable Diffusion 3 API delivers powerful image generation
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.
Explore other upcoming enterprise technology events and webinars powered by TechForge here.
SAS has added expanded capabilities to its SAS Viya flagship data and AI platform, including the general availability of SAS Viya Workbench.
Targeted to developers and modellers, Viya Workbench is a self-service, on-demand compute environment for conducting data preparation, exploratory data analysis, and developing analytical and machine learning models.
Kathy Lange, Research Director for AI Software, IDC, said: “While SAS Viya is the foundation of the SAS software ecosystem, the company is evolving its portfolio with innovative products to meet diverse user needs.
“New offerings like SAS Viya Workbench – an on-demand analytics environment – aim to increase productivity, enhance performance and build trust among AI developers.”
Developer canvas for experimentation and exploration
Viya Workbench allows developers and modellers to work in the language of their choice, initially SAS and Python, with R available by the end of 2024. Using an intuitive and flexible interface, Viya Workbench offers two development environment options – Jupyter Notebook/JupyterLab and Visual Studio Code.
Tapping into powerful SAS analytical procedures (PROCs) and native Python APIs within Viya Workbench accelerates development of high-performance AI models. Additionally, state-of-the-art, custom Python libraries – unique to Viya Workbench – can significantly improve speed and performance with minimal changes to a developer’s existing Python program.
Viya Workbench is a flexible, scalable and efficient development environment that is on-demand, self provisioning and self-terminating with minimal IT support. The dedicated analytical environment features customisable CPU/GPU compute power to match the needs of the project. Models and other results can be leveraged in SAS Viya for data management, governance and operational deployment.
Viya Workbench will initially be available through the Amazon AWS Marketplace in Q2, with future plans for additional supported cloud providers and a software-as-a-service deployment option.
Increased developer productivity, faster AI innovation
The business case for Viya Workbench is strong. AI developers and modellers want to work with modern, open source packages and cutting-edge cloud compute, but they are also under pressure to deliver fast results and manage costs. In addition, they want prebuilt, scalable infrastructures that allow them to focus on creating, innovating, iterating and testing their work.
Jared Peterson, Senior Vice President of Engineering, SAS, said: “The many challenges developers face aren’t just minor annoyances – they are obstacles that prevent questions from being answered and work from getting done.