Arm, the British semiconductor and software design powerhouse, staged its Arm Everywhere China showcase in Shanghai to announce a suite of new AI‑driven technologies aimed at reshaping mobile graphics. The centerpiece of the announcement is a full AI‑native computing ecosystem that includes the CSS for Mobile 2 compute platform and the freshly unveiled Arm Mali G2‑Ultra NX GPU. This GPU is built around neural processing units and promises to deliver higher graphics performance while dramatically cutting heat output and power draw compared with traditional designs. Arm positions its on‑device AI solution as a more transparent alternative to Nvidia’s controversial DLSS 5.
Rather than relying on a closed‑source, generative model, Arm’s approach uses convolution‑based networks that are trained on real‑world content and released openly on platforms such as GitHub and Hugging Face. This openness, according to Arm’s GPU product lead Deyan Lazarov, gives developers the ability to fine‑tune the models to match the unique visual style of their games, preserving artistic intent and intellectual‑property control. While Arm has not disclosed the exact devices that will ship with the Mali G2‑Ultra NX, the predecessor Mali G1‑Ultra already powers phones from Xiaomi, Vivo and OnePlus, suggesting a similar tier of flagship or high‑end mid‑range smartphones could receive the new chip. Lazarov explained that the team has spent roughly two and a half years developing the GPU, with a clear focus on enabling compute‑heavy features such as ray tracing on mobile hardware.
"There was a growing adoption of ray tracing in mobile titles," he noted, "the technology was slow to take off, but now more games are actively looking to enable it." The new GPU integrates tightly with Unreal Engine, allowing developers to leverage advanced features like MegaLights. Arm partnered with Sumo Digital to create a demo title, Neural Dawn, that showcases the chip’s capabilities. Other announced partners include Tencent’s Arena Breakout Infinite, NetEase’s Where Winds Meet, and Infold Games’ Infinity Nikki, all of which will benefit from the AI‑enhanced rendering pipeline.
One of the core challenges of mobile gaming is the limited thermal envelope and battery‑driven power budget, which differ sharply from consoles or PCs. Lazarov explained that the Mali G2‑Ultra NX tackles this by embedding neural processing directly into the rendering pipeline. The AI engine can reconstruct large portions of each frame, reducing the amount of work the GPU must perform and consequently lowering power consumption. This efficiency gain translates into more headroom for developers, who can choose to reinvest the saved cycles into higher visual fidelity, smoother frame rates, or longer battery life.
Arm’s suite of AI‑based rendering techniques includes Neural Super Sampling (NSS), Neural Super Sampling and Denoising (NSSD), and Neural Frame Rate Upscaling (NFRU). In practice, NSS can render only a quarter of an image on the GPU and have the neural accelerator fill in the remaining three‑quarters. In an even more extreme scenario, the entire frame can be generated by the neural hardware with no traditional rasterization at all.
This contrasts with Nvidia’s DLSS 5, which uses generative networks that can introduce artifacts or deviate from a developer’s artistic vision. Arm’s convolutional networks, by contrast, simply learn how existing content should look and then apply that knowledge without “imagining” new details. The openness of Arm’s models is a key differentiator. All of the resolution‑upscaling and frame‑generation networks are publicly available, allowing developers to retrain them on custom datasets that reflect the unique aesthetic of their games.
Lazarov emphasized that this eliminates the risk of losing IP control, a concern with proprietary black‑box solutions. Developers can download the models, feed them game‑specific imagery, and produce a version of the network that perfectly matches their style. From a performance standpoint, Arm claims the Mali G2‑Ultra NX delivers up to four times the performance and efficiency of its predecessor, along with a 70 % reduction in DRAM traffic—a major source of power draw on mobile devices.
In benchmark tests, the older GPU could only push Neural Dawn to about 40 fps at its limit and dropped to roughly 15 fps under typical conditions. The new chip, however, comfortably maintains 60 fps, illustrating the practical impact of the AI‑assisted pipeline. Arm is also investing heavily in developer tooling. Plug‑ins for Unreal Engine are already available, and a Unity integration is on the roadmap, complemented by a recent partnership with Unity China.
For studios using custom engines, Arm provides a comprehensive SDK that streamlines integration without requiring engine‑specific plug‑ins. Lazarov noted that even legacy titles can be upgraded with Arm’s AI graphics in as little as three months, giving developers a fast path to modern visual quality. Overall, Arm’s mobile‑first AI reconstruction strategy aims to give developers a transparent, flexible, and power‑efficient alternative to existing solutions.
By making the underlying networks open, optimizing for the strict thermal and battery constraints of phones, and delivering tools that reduce the workload on traditional GPUs, Arm hopes to democratize high‑end graphics on mobile platforms and bring a desktop‑like experience to handheld devices.