Arm, the UK‑based semiconductor and software design powerhouse, recently hosted its Arm Everywhere China showcase in Shanghai, using the stage to announce a suite of fresh AI‑computing solutions aimed at the mobile market. The centerpiece of the announcement is a new AI‑native compute platform, CSS for Mobile 2, paired with the upcoming Mali G2‑Ultra NX GPU – a graphics processor built around neural‑accelerator technology that promises higher visual fidelity while keeping power draw and heat generation modest. Arm positions this on‑device AI‑enhanced graphics pipeline as a more transparent and developer‑friendly alternative to Nvidia’s controversial DLSS 5, arguing that its method aligns better with the original intent of game creators.
While Arm did not disclose specific handset models that will ship with the Mali G2‑Ultra NX, the predecessor Mali G1‑Ultra has already been integrated into smartphones from brands such as Xiaomi, Vivo and OnePlus, suggesting a similar market trajectory for the new chip. 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 under development for roughly two and a half years. Its primary goal is to give mobile developers access to compute‑heavy features—most notably real‑time ray tracing—without sacrificing the battery life or thermal envelope that mobile devices must respect.
"Ray tracing is finally gaining traction on phones," Lazarov noted. "It moved slowly at first, but we’re now seeing a growing number of titles that want to enable it." The GPU’s capabilities have already been demonstrated in a partnership with Sumo Digital, which built a showcase title called Neural Dawn using Unreal Engine’s MegaLights technology. This demo highlights how the Mali G2‑Ultra NX can deliver console‑like visual quality on a handheld form factor.
Lazarov described the target audience for the chip as "mobile‑first games that aspire to look like desktop experiences," as well as cross‑platform titles that need to deliver a consistent visual level whether they run on a console, a PC, or a phone. He emphasized the practical challenges of mobile hardware: unlike consoles, phones cannot dissipate large amounts of heat and must run continuously on battery power. "A phone’s thermal budget and power envelope are fundamentally different from a PC or a console," he said. "You can’t just dump heat the way a desktop can." Arm’s answer is to weave neural processing directly into the traditional rendering pipeline.
By offloading portions of the image generation to dedicated neuro‑accelerators, the GPU workload is reduced, which in turn cuts power consumption and improves overall system efficiency. "That extra efficiency means developers don’t have to make as many trade‑offs between visual fidelity, frame rate, and battery life," Lazarov explained. "They can choose where to spend the saved budget—whether that’s higher resolution, richer effects, smoother performance, or longer playtime." To make the technology attractive to creators, Arm is taking an open‑source stance.
Over the past two years the company has worked closely with the broader ecosystem to provide tools that are both powerful and accessible. The core AI models—Neural Super Sampling (NSS), Neural Super Sampling and Denoising (NSSD), and Neural Frame Rate Upscaling (NFRU)—are released publicly on GitHub and Hugging Face. Unlike Nvidia’s DLSS, which relies on proprietary, generative networks that act as a black box, Arm’s convolution‑based networks do not invent new content; they simply recognize patterns in real‑world data and reconstruct missing information accordingly. Lazarov illustrated the process with a two‑frame example.
In the first frame, the GPU renders only a quarter of the final image; the remaining three‑quarters are filled in by the neuro‑accelerators using NSS. In the second frame, the entire image is produced by the AI engines with no traditional GPU rendering at all, demonstrating a fully AI‑reconstructed frame. This approach enables up to four‑times performance gains and similar efficiency improvements, while also slashing DRAM traffic by roughly 70 %—a major source of power draw on mobile devices. Arm’s open model also gives developers the ability to fine‑tune the networks for their own artistic styles.
Because the training data and the model code are public, a studio can capture screenshots from its own game, retrain the network on that dataset, and ensure that upscaled or interpolated frames retain the unique visual identity of the title. This mitigates the risk of losing intellectual‑property control that can occur with closed‑source solutions. In contrast, DLSS 5 is described by Lazarov as a "take‑it‑or‑leave‑it" proposition: a proprietary system that cannot be retrained or inspected, and which would simply exceed the power and bandwidth limits of a smartphone even if the software were ported. Arm’s solution, by contrast, was engineered from the ground up with mobile constraints in mind, balancing power, memory bandwidth and thermal headroom.
Real‑world performance numbers back up the claims. When pushed to its limits, the previous‑generation Mali GPU could only run Neural Dawn at about 40 fps under ideal conditions and fell to roughly 15 fps in typical gameplay. The Mali G2‑Ultra NX, however, comfortably sustains 60 fps, delivering smoother animation and a more responsive experience. On the developer side, Arm currently offers Unreal Engine plug‑ins and a roadmap that includes Unity support, acknowledging that many studios use a variety of engines.
An SDK is also available for custom engines, providing a streamlined integration path. Lazarov highlighted that retrofitting an existing shipped title with Arm’s AI‑native graphics pipeline can be accomplished in as little as three months, enabling rapid upgrades without a full rebuild. The ecosystem is already showing results.
Titles such as Tencent’s Arena Breakout Infinite, NetEase’s Where Winds Meet, and Infold Games’ Infinity Nikki are slated to leverage the new GPU. Moreover, Arm’s partnership with Unity China signals a broader commitment to reach developers wherever they work.
In summary, Arm’s mobile‑first AI reconstruction technology offers a transparent, open, and power‑efficient alternative to Nvidia’s DLSS 5. By embedding neural accelerators into the rendering flow, providing open‑source models that can be customized, and delivering concrete performance gains, Arm aims to empower developers to create visually stunning games that run smoothly on battery‑powered devices without compromising artistic intent or hardware limitations.