Arm filed 2 patents in H1 2026, all in Graphics.
The filings focus on neural frame interpolation technology that generates intermediate frames for real-time Graphics applications using learned alpha blending techniques. This approach aims to increase perceived frame rates on mobile and low-power devices without requiring full frame rendering, addressing power budget constraints in smartphones and tablets during gaming.
Both Graphics patents filed in this period center on the same core challenge: generating smoother frame rates on mobile devices without the computational cost of rendering every frame from scratch. One application describes a neural network that learns optimal alpha blending parameters to produce intermediate frames, rather than relying on fixed interpolation formulas, allowing the output frames to be more accurate while keeping the processing load low enough for mobile GPUs. The other takes a pipeline-level approach, combining neural frame rate upsampling with a design tuned specifically to the thermal and bandwidth limits of smartphone and tablet Hardware, addressing a gap left by existing solutions that were built around the power budgets of desktop and console GPUs.
All data sourced from USPTO patent filings. Google Patents may take several weeks to index recent publications. If a link is unavailable, search for the patent number at USPTO Patent Public Search.