PC and console players have grown used to AI graphics upscaling over the past several years. Now ARM Holdings, one of the biggest names in mobile chips, is bringing similar technology to smartphones and tablets, opening a new chapter in mobile graphics.
Neural Super Sampling: Better-Looking Graphics at a Lighter Load for Mobile GPUs
ARM has just introduced a new technology called Neural Super Sampling (NSS), an AI-based, real-time image upscaling method that doubles resolution at a cost of only about 4 milliseconds of extra processing per frame. NSS lets game developers render at a resolution as low as 540p while outputting an image that comes close to 1080p, significantly reducing the workload on the GPU.
According to ARM, using NSS can cut GPU load by up to 50%, which saves power or allows noticeably higher frame rates (FPS) on mobile devices. It is a major step forward in optimizing graphics performance on smartphones, especially in visually demanding games.
On the PC side, Nvidia led the way with Deep Learning Super Sampling (DLSS) back in 2019 on the GeForce RTX line. ARM's new technology appears comparable to DLSS 1.0 or 2.0, but it is designed specifically for the mobile device ecosystem.
https://www.youtube.com/watch?v=XmdsWErzwC0
NSS also builds on Accuracy Super Resolution, ARM's earlier solution based on AMD's FidelityFX Super Resolution 2 (FSR2) technology. FSR2 is already integrated into several prominent games, such as Fortnite and Infinity Nikki.
ARM's Development Tools and Expansion Roadmap
To help developers integrate NSS more easily, ARM has announced the Neural Graphics Development Kit, a development toolset built specifically for mobile. The package includes:
- A plugin for Unreal Engine,
- A Vulkan emulator for PC,
- A profiling tool for performance optimization,
- ML extensions for the Vulkan graphics API.
ARM isn't stopping there. It has also decided to open-source the AI models behind its graphics upscaling technology on GitHub and Hugging Face. This makes them easier for the developer community to access and speeds up research into practical applications.
Beyond NSS, ARM is also planning to roll out other AI technologies, such as:
- Neural Frame Rate Upscaling (boosting frame rates),
- Denoising for path tracing (advanced removal of image noise),
- And other variants of Neural Super Sampling.
These technologies are expected to reach developers starting in 2026, before hardware with full support launches.
ARM's heavy investment in AI graphics for mobile points to a shift in the gaming and imaging industry, as users increasingly demand a more premium experience right on their smartphones. If it succeeds, ARM could help narrow the gap between PC and mobile graphics, producing a new generation of mobile games that look sharper and run smoother than ever.
