Arm, the British chip and software design powerhouse, staged its Arm Everywhere China showcase in Shanghai to announce a suite of fresh AI‑computing innovations aimed squarely at mobile devices. The centerpiece of the announcement is a new AI‑native compute platform called CSS for Mobile 2, paired with the Mali G2‑Ultra NX graphics processor – a GPU built around dedicated neural accelerators that promises higher visual fidelity while keeping power draw and heat generation to a minimum. Arm positions this on‑device AI pipeline as an open, developer‑friendly alternative to Nvidia's contentious DLSS 5, arguing that its method stays truer to the original artistic intent of game creators. While Arm did not disclose the exact products that will ship with the Mali G2‑Ultra NX, the predecessor Mali G1‑Ultra already powers smartphones from brands such as Xiaomi, Vivo and OnePlus, suggesting that the new chip will soon appear in a similar range of high‑end handsets.
Deyan Lazarov, a senior member of Arm’s GPU product team within the Edge AI division, explained that the Mali G2‑Ultra NX has been in development for roughly two and a half years. Its primary goal is to give mobile developers access to heavyweight rendering techniques—most notably ray tracing—without the prohibitive thermal and battery penalties that have traditionally limited such features on phones and tablets. "There’s a clear upward trend in mobile titles adopting ray tracing," Lazarov said. "It was a slow burn at first, but now we see more developers eager to integrate realistic lighting and reflections into their games." The new GPU, according to a GamesIndustry.biz report earlier this year, supports Unreal Engine features like MegaLights, and Arm has already partnered with Sumo Digital to produce a demo title called Neural Dawn that showcases the chip’s capabilities.
Lazarov emphasized that the Mali G2‑Ultra NX is built for "mobile‑first" experiences that still aspire to look like desktop‑class games. He described three categories of titles the chip targets: (1) games designed primarily for phones that want to achieve console‑level visual quality; (2) cross‑platform titles that run on consoles, PCs and mobiles and therefore need a consistent look across devices; and (3) on‑the‑go experiences where a player might switch from a home console to a train seat without a noticeable drop in graphics fidelity. Among the announced partners are Tencent’s Arena Breakout Infinite, NetEase’s Where Winds Meet and Infold Games’ Infinity Nikki.
One of the core challenges for mobile graphics is the limited thermal envelope and battery capacity. "A phone can’t dump heat the way a console or a desktop can," Lazarov noted.
"It runs on a battery all day, so every watt matters." Arm’s answer is to blend neural processing directly into the rendering pipeline. By offloading portions of the image reconstruction to dedicated neural accelerators, the GPU workload shrinks, power consumption falls, and the overall system becomes more efficient.
This efficiency, Lazarov argued, frees developers from many of the compromises they normally make between image quality, frame rate and battery life. To make the technology attractive, Arm is taking an open‑source stance. Over the past two years the company has been collaborating with the wider ecosystem to produce AI tools that are both powerful and accessible. "We wanted an AI solution that developers could actually understand and tweak," Lazarov explained.
"Our networks are convolution‑type, not generative, so they don’t hallucinate new content—they simply know how to enhance what’s already there based on real‑world training data. And because we publish the models on GitHub and Hugging Face, anyone can download, fine‑tune or even retrain them for a specific art style." Arm’s suite includes Neural Super Sampling (NSS), Neural Super Sampling and Denoising (NSSD) and Neural Frame Rate Upscaling (NFRU).
In practice, a frame can be rendered at only a quarter of its native resolution, with the neural accelerator reconstructing the missing three‑quarters in real time. In more extreme cases, the entire frame can be generated by the AI engine without any traditional rasterisation, dramatically cutting GPU cycles. Lazarov contrasted this with Nvidia’s DLSS 5, which relies on proprietary, closed‑source generative models.
"DLSS is a black box—you can’t see what it was trained on or adjust it for your unique visual language," he said. "Our approach is transparent, modifiable, and designed to respect the developer’s IP." The mobile‑first philosophy also informs performance targets. Arm claims the Mali G2‑Ultra NX can deliver up to four times the performance and efficiency of its predecessor while slashing DRAM traffic by roughly 70 percent—a major win because external memory accesses are a leading source of power drain on smartphones. In a head‑to‑head test, the previous‑generation GPU managed only about 40 fps on Neural Dawn under ideal conditions and dropped to 15 fps in typical usage.
The new chip, by contrast, comfortably sustains 60 fps, delivering a smoother experience without overheating. From a developer tooling perspective, Arm currently offers plug‑ins for Unreal Engine and is working toward Unity support, having recently announced a partnership with Unity China.
For studios that rely on custom engines, Arm provides an SDK that abstracts the low‑level integration steps, enabling a relatively quick path to AI‑enhanced graphics. Lazarov estimates that retrofitting an existing shipped title with Arm’s AI pipeline can take as little as three months, a timeline that could dramatically shorten the lifecycle of post‑launch updates.
The demo game Neural Dawn, built by a team of fewer than 20 people in under 18 months, serves as a proof‑of‑concept for how Arm’s technology can accelerate development. By leveraging the AI‑native graphics stack, developers can either boost visual fidelity, improve frame rates, extend battery life, or strike a balance among these goals without rewriting large portions of their codebase. In summary, Arm’s mobile‑first AI reconstruction strategy seeks to democratize high‑quality graphics on handheld devices through open, efficient neural networks that complement traditional rendering. By addressing power, thermal and memory constraints head‑on, and by giving developers the freedom to tailor the AI models to their artistic vision, Arm aims to create a more flexible and transparent alternative to Nvidia’s closed‑source DLSS ecosystem, potentially reshaping how next‑generation mobile games look and feel.