Arm Details Its Mobile-First AI Reconstruction Approach, Diverging from the "Black Box" Model of DLSS 5

Arm, the UK‑based semiconductor and software design powerhouse, staged its Arm Everywhere China showcase in Shanghai to announce a suite of fresh AI‑computing innovations aimed at the mobile market. Central to the reveal is a comprehensive AI‑native ecosystem that includes the next‑generation compute platform CSS for Mobile 2 and the brand‑new Arm Mali G2‑Ultra NX graphics processor. This GPU is built around neural‑accelerated rendering, promising a leap in visual fidelity while keeping power draw and heat generation modest.

Arm positions its on‑device AI solution as a more transparent alternative to Nvidia’s contentious DLSS 5, aligning more closely with developers’ original creative intent. While Arm did not disclose specific product partners for the Mali G2‑Ultra NX, its predecessor, the Mali G1‑Ultra, already powers devices from Xiaomi, Vivo and OnePlus, suggesting a similar rollout path. Deyan Lazarov, a senior engineer on Arm’s GPU team within the Edge AI division, explained that the chip has been under development for roughly two and a half years. Its design goal is to give mobile developers access to heavyweight graphics techniques—most notably ray tracing—without sacrificing the battery life or thermal envelope of a smartphone.

"We’re seeing a growing appetite for ray tracing on mobile titles," Lazarov noted. "It was a slow burn at first, but more games are now asking for that realistic lighting and reflection capability." Recent reports from GamesIndustry.biz highlighted that the new GPU will let developers tap into Unreal Engine features such as MegaLights. Arm even collaborated with Sumo Digital to produce the demo title Neural Dawn, which showcases the Mali G2‑Ultra NX’s capabilities. Lazarov described the target audience as "mobile‑first games that still want desktop‑grade visuals".

He added that many titles are built to be cross‑platform, allowing players to enjoy a seamless experience whether they’re on a home console, a laptop, or a phone on a commuter train. The challenge, he explained, is that phones operate under far tighter thermal and power constraints than consoles or PCs.

"A phone can’t dump heat the way a console can, and it’s always running off a battery," he said. Arm’s answer is to fuse neural processing with the conventional rendering pipeline. By offloading portions of the workload to dedicated neuro‑accelerators, the GPU can render fewer native pixels while the AI reconstructs the rest, slashing both compute demand and energy consumption.

This efficiency gain translates into more headroom for developers, reducing the need for hard compromises between visual quality, frame rate, and battery life. To foster adoption, Arm is taking an open‑source stance. Over the past two years the company has built a toolbox of software resources that developers can integrate into their pipelines. "We wanted AI that is developer‑friendly and that delivers tangible efficiency gains," Lazarov explained.

"Those gains can be reinvested into sharper graphics, higher performance, longer battery life, or any mix of the three." Key components of Arm’s suite include Neural Super Sampling (NSS), Neural Super Sampling and Denoising (NSSD), and Neural Frame Rate Upscaling (NFRU). In practice, NSS can upscale a low‑resolution render by reconstructing missing detail, while NFRU generates additional frames to smooth motion. Lazarov illustrated the process with a two‑frame example: in the first, the GPU renders only 25 % of the image and the neuro‑accelerators fill in the remaining three‑quarters via NSS; in the second, the entire frame is produced by the AI, with the GPU doing virtually no traditional rendering work.

Nvidia’s DLSS 5 also uses AI to enhance images, but it has drawn criticism for its "black‑box" nature and for sometimes diverging from a developer’s artistic vision. Arm draws a sharp contrast, emphasizing that its networks are convolutional rather than generative. "Our networks don’t imagine new content; they’ve been trained on real data to recognize how things should look," Lazarov said. "All of our models—resolution upscalers, frame‑rate generators—are publicly available on GitHub and Hugging Face." This openness lets developers fine‑tune the models on their own art assets.

If a game has a distinctive visual style, a studio can capture data from that game, retrain the open model, and ensure the AI upscaling respects the intended aesthetic. This eliminates concerns about IP loss or unwanted stylistic changes, a risk that developers associate with proprietary solutions like DLSS. Lazarov also highlighted the mobile‑first philosophy baked into Arm’s design. The neural networks are built to respect the strict power, bandwidth, and memory budgets of smartphones.

By contrast, DLSS would quickly exceed those limits on a phone, even if the software were ported. Arm reports up to four‑times performance and efficiency improvements, along with a 70 % reduction in DRAM traffic—a major source of power draw on mobile devices.

Performance benchmarks illustrate the leap. Running Neural Dawn on the previous‑generation Mali GPU capped at about 40 fps under ideal conditions and fell to 15 fps in real‑world scenarios.

The Mali G2‑Ultra NX, however, comfortably sustains 60 fps, delivering a fluid experience without overheating. On the tooling side, Arm currently offers plug‑ins for Unreal Engine and plans to expand to Unity and other engines. A recent partnership with Unity China signals that Unity support is on the roadmap, though Unreal remains the primary focus for now. For studios using custom engines, Arm provides an SDK that streamlines integration without requiring a dedicated plug‑in.

Lazarov assures developers that retrofitting existing titles with Arm’s AI‑native graphics can be swift. "If you have a shipped game and want to add our solution, you could have a new version ready in as little as three months," he said. This rapid turnaround is exemplified by Neural Dawn itself, which a sub‑20‑person team delivered in under 18 months.

In summary, Arm’s mobile‑first AI reconstruction strategy offers a transparent, open, and power‑efficient alternative to Nvidia’s DLSS 5. By embedding neural upscaling and frame generation directly into the graphics pipeline, Arm enables developers to push visual fidelity on constrained devices while preserving battery life and thermal comfort.

The Mali G2‑Ultra NX, backed by a growing ecosystem of tools and open models, aims to bring desktop‑level graphics to the palm of your hand, reshaping what mobile gaming can look like in the years ahead.