ARM’s new “Neural Super Sampling” (NSS) technology is a frame-level AI upscaler designed specifically for the constraints of smartphone hardware. The neural upscaler takes low-resolution frames—such as 540p—and uses a trained AI model to reconstruct them at higher fidelity, often targeting 1080p output, while substantially reducing the performance overhead compared to native rendering. By leveraging temporal data from multiple frames, ARM’s NSS produces visuals with minimal edge softening or texture muddiness, matching or closely approximating native resolution in both static and motion scenes.
As a major ARM licensee, Qualcomm is positioned to integrate Neural Super Sampling into its future Snapdragon system-on-chips, tapping into the on-die neural processing units (NPUs) that already deliver up to 45 TOPS of AI throughput. Embedding NSS directly into Snapdragon platforms would enable game developers to offload super-sampling tasks to dedicated AI hardware, boosting framerates on AAA mobile titles without compromising visual quality or exceeding typical power budgets. This collaboration between ARM’s neural upscaler and Qualcomm’s AI-capable NPUs could redefine on-device gaming performance, reducing reliance on cloud rendering and discrete GPU power.
Image Credit: ARM
What's Driving This Trend
- AI-driven Mobile Graphics
- The integration of AI-driven graphics upscaling like Neural Super Sampling into mobile devices heralds a new era of portable gaming with console-level visuals.
- Efficient Edge AI Processing
- Advancements in edge AI processing enable sophisticated tasks such as graphics upscaling to be performed on-device, reducing the need for constant connectivity to cloud services.
- On-device Computational Improvements
- The continued enhancement of on-device computational capabilities, through innovations like advanced neural processors, is transforming how graphics are processed on smartphones.
Who This Affects Most
- Mobile Gaming
- The mobile gaming industry stands to gain immensely as AI-powered graphics upscaling technology boosts performance and visual fidelity without increasing energy consumption.
- Semiconductors
- The semiconductor industry is evolving rapidly as companies like ARM and Qualcomm integrate advanced AI capabilities directly into chip designs, enhancing processing efficiency.
- Augmented Reality (AR)
- Developments in hardware upscaling technologies open up new possibilities for augmented reality applications, allowing for richer, more immersive experiences on mobile devices.
