adminUnreal Engine MegaLights on mobile could make advanced dynamic lighting and ray tracing more practical for handheld gaming. Arm and UK studio Sumo Digital are demonstrating the approach in Neural Dawn, a mobile game built around Arm’s neural graphics technology and Mali GPU hardware.
How Unreal Engine MegaLights on mobile works
Neural Dawn uses neural processing alongside the GPU to handle demanding graphics tasks. Arm says this close coordination can reduce the time spent moving work between graphics and machine-learning processes, while helping developers target detailed visuals without using power unnecessarily.
The game is designed to showcase two technologies: Neural Super Sampling and Denoising (NSSD) and Neural Frame Rate Upscaling (NFRU). Together, they address image quality and frame rate, two important considerations for mobile games. For players, the practical benefit depends on the device, the game’s settings and how efficiently the hardware manages heat and battery use.
Neural Dawn features dynamic lighting. | Image credit: Sumo Digital/Arm
NSSD reduces ray-tracing noise
Ray tracing can require substantial processing, so NSSD is designed to produce a clean image from fewer rays. Using less ray data can leave gaps or visible noise in a scene; the denoising process reconstructs the image using information such as colours and shapes.
Arm’s Peter Hodges describes the method as image reconstruction rather than image generation. The distinction matters because the goal is to restore detail from incomplete rendering information, not invent content that was not present. Arm has also published the models and says developers can inspect and retrain them for their own game content.
Practical takeaway: fewer rays may lower the rendering workload, but developers still need to test image quality across different scenes and device conditions.
NFRU can increase the displayed frame rate
Neural Frame Rate Upscaling takes two consecutive frames and creates an intermediate frame between them. In the example described for Neural Dawn, this can raise 30 frames per second content to 60 fps. The result is intended to make motion appear smoother, although the final experience can vary with the game and hardware.
What the technology means for game development
Sumo Digital’s Gary Dunn said the collaboration enabled the team to use MegaLights and ray tracing while saving power, potentially supporting longer play sessions. Art director Lukáš Medek said the performance headroom allowed the team to approach asset counts, texture sizes, materials and lighting more like a PC or console project.
The team of 17 developed the roughly two-hour game over 18 months. Its story follows a research scientist exploring a cave network guided by light. Because the lighting is dynamic, the developers could work without relying on the traditional process of baking light maps, which can make lighting changes slower to iterate.
Arm says the Mali GPU technology will deliver visuals comparable to desktop-class experiences while maintaining battery life. That is a stated goal, not a guarantee for every phone: performance and battery consumption depend on the device and the game’s implementation.
Release details for Neural Dawn
Neural Dawn is scheduled to launch later in 2026, exclusively on Android devices powered by Arm Mali GPUs. Arm’s technology is intended to demonstrate what neural graphics could enable on compatible mobile hardware, but the game’s device requirements and availability should be checked closer to release.
Overall, Unreal Engine MegaLights on mobile represents a push toward more advanced real-time lighting on phones. NSSD aims to make ray tracing more manageable, while NFRU targets smoother motion. For players, the key details to watch are supported devices, visual quality and how well performance holds up during longer sessions.
