Neural Super Sampling (NSS), a next-generation AI-powered upscaling solution from Arm is released for graphics and gaming developers to start out experimenting today!
Elevated by machine learning
NSS is designed for real-time performance on future mobile devices with Arm Neural Technology. Nevertheless, latency is determined by implementation aspects reminiscent of GPU configuration, resolution, and use case. In our Enchanted Castle demo video below, NSS reduced GPU workload by 50 percent. The model rendered at 540p and upscaled to 1080p in only 4ms in sustained performance setup.
Find out about our NSS Model
Neural Super Sampling (NSS) is a parameter prediction model for real-time temporal super sampling developed by Arm, optimized for execution on Neural Accelerators (NX) in mobile GPUs. It enables high-resolution rendering at a lower compute cost by reconstructing high-quality output frames from low-resolution temporal inputs. NSS is especially fitted to mobile gaming, XR, and other power-constrained graphics use cases.
Start with our NSS model today.
If you wish to go deeper try the next resources:
How we trained the model
Try the Neural Graphics Dataset: A set of reference images and image sequences together with the corresponding motion, depth and other data required to coach, validate and test neural super sampling algorithms.
The present version of the dataset features a limited set of information for Neural Super Sampling to display how the NSS model development flow works. While this flow doesn’t yet provide a comprehensive dataset for complete model (re)training, stay tuned for future releases of the Neural Graphics Model Gym where Arm will provide tools to capture and convert content to be used in model training and retraining.
Start experimenting with NSS today!
NSS has been integrated into Unreal Engine via two plugins, the NSS Plugin for Unreal® Engine and Unreal® NNE Plugin for ML extensions for Vulkan.
For step-by-step instructions on the right way to use NSS in Unreal® Engine.See our learning paths:
