Arm, the British chip and software design powerhouse, recently hosted its Arm Everywhere China showcase in Shanghai, where it introduced a suite of new AI‑driven technologies aimed at reshaping graphics on mobile devices. The centerpiece of the announcement was the AI‑native compute platform CSS for Mobile 2 and the Mali G2‑Ultra NX GPU, a processor built around neural‑accelerated rendering.
Arm says this combination can deliver desktop‑class visual quality on phones, tablets and other battery‑powered gadgets while keeping heat and power consumption in check. While the company did not name specific products that will ship with the Mali G2‑Ultra NX, the predecessor Mali G1‑Ultra is already present in smartphones from brands such as Xiaomi, Vivo and OnePlus, suggesting that the new chip could appear in a similar range of consumer devices. Deyan Lazarov, a senior member of Arm’s GPU product team within the Edge AI division, explained that development on the Mali G2‑Ultra NX has spanned roughly two and a half years.
The goal, he said, was to give mobile developers the ability to harness compute‑heavy features—most notably real‑time ray tracing—without sacrificing the battery life or thermal envelope that mobile platforms demand. "We’ve seen a steady rise in interest for ray tracing on phones," Lazarov noted. "It started out slow, but now more titles are actively looking to integrate it." The new GPU, according to Arm, works hand‑in‑hand with Unreal Engine’s advanced lighting tools such as MegaLights. In a joint effort with Sumo Digital, Arm demonstrated the capabilities of the chip through a demo game called Neural Dawn, which showcases AI‑enhanced graphics while running smoothly on a mobile form factor.
Lazarov described the target audience for the Mali G2‑Ultra NX as developers who adopt a "mobile‑first" mindset but still aspire to the visual fidelity of desktop games. He added that many modern titles are designed to be cross‑platform from the outset, meaning players expect a consistent experience whether they are on a console at home or a smartphone on a train.
To meet those expectations, the GPU must overcome the unique constraints of mobile hardware—limited thermal headroom, strict power budgets and reliance on battery power. "A phone can’t dump heat the way a console can, and it can’t draw unlimited power," Lazarov explained. "Our solution blends neural processing with the traditional rendering pipeline, trimming the GPU’s workload and slashing its power draw. That translates into system‑level efficiency gains and gives developers more flexibility when balancing visual quality, frame rate and battery life." Arm is backing this hardware with an open‑source software stack and a collection of developer tools.
Lazarov emphasized that the company has spent the past two years collaborating with the broader ecosystem to make AI‑driven graphics approachable. "We wanted a method that’s developer‑friendly, letting creators reap huge efficiency gains and then decide how to reinvest those savings—whether that’s into sharper textures, higher frame rates, longer battery life, or a mix of all three." Key components of Arm’s AI suite include Neural Super Sampling (NSS), Neural Super Sampling and Denoising (NSSD) and Neural Frame‑Rate Upscaling (NFRU). These techniques enable the system to upscale lower‑resolution frames, synthesize additional rays for better lighting, and generate extra frames using AI‑based reconstruction.
Lazarov illustrated the process with a two‑frame example: in the first, the GPU renders only a quarter of the image while the neuro‑accelerators fill in the remaining three‑quarters via NSS. In the second frame, the GPU does no traditional rendering at all; every pixel is produced by the neural engines, demonstrating a fully AI‑reconstructed output.
Nvidia’s competing DLSS 5 technology also uses AI to enhance image detail, but it has drawn criticism for its "black‑box" nature and the risk of deviating from a developer’s artistic intent. Arm draws a clear line between its approach and Nvidia’s.
"Our networks are convolutional, not generative," Lazarov said. "They don’t imagine new content; they simply learn how existing content should look.
Because we train them on real data and publish the models on GitHub and Hugging Face, developers can inspect, fine‑tune or completely retrain them to match a unique art style. This preserves IP and ensures the upscaling respects the creator’s vision." By contrast, DLSS is a proprietary, closed system where developers have no visibility into the training data and cannot modify the model. "It’s a take‑it‑or‑leave‑it proposition," Lazarov remarked.
"Our solution is take‑it‑and‑make‑it‑your‑own." Another distinction is the mobile‑first design philosophy behind Arm’s tech. The neural networks are built to operate within the tight power, bandwidth and memory constraints of smartphones. "If you tried to run DLSS on a phone, you’d instantly blow past the budget," Lazarov warned. Power efficiency is a major selling point.
Arm claims the Mali G2‑Ultra NX delivers up to four times the performance and efficiency of its predecessor, while slashing DRAM traffic by roughly 70 %—a critical factor because external memory accesses are a major source of power consumption on mobile chips. In practical terms, the new GPU can run Neural Dawn at a steady 60 fps, whereas the older generation struggled to reach 40 fps under ideal conditions and fell to 15 fps in typical use. On the tooling side, Arm currently offers plugins for Unreal Engine and has plans to expand to Unity and other engines.
A recent partnership with Unity China signals that Unity support is on the roadmap, even though Unreal remains the immediate focus. For developers using custom engines, Arm provides an SDK that streamlines integration without requiring a dedicated plugin.
Lazarov highlighted that existing games can be upgraded with Arm’s AI graphics in as little as three months, allowing studios to deliver visual improvements to shipped titles quickly. He cited Neural Dawn’s development timeline—fewer than 20 engineers built the demo in under 18 months—as evidence of how Arm’s technology accelerates production while keeping resource demands modest. In summary, Arm’s mobile‑first AI reconstruction strategy offers an open, customizable alternative to Nvidia’s closed DLSS 5, promising substantial gains in visual fidelity, frame‑rate and power efficiency for a new generation of mobile gaming experiences.