Nvidia Announces MotionBricks, a Real-Time Motion Generation Framework
MotionBricks could simplify the complex animation graphs used in large-scale games and robotics by generating motions from keyframe constraints without retraining for each task.
Reporting from 1 sources: GIGAZINE.
Nvidia's AI research team announced MotionBricks, a motion generation framework for game animation and robot control. It handles over 350,000 motion skills with a single neural backbone, achieving 15,000 FPS and 2 ms latency. The technology was presented as a SIGGRAPH 2026 paper accepted by ACM Transactions on Graphics.
MotionBricks uses operational units called 'smart primitives' to absorb the complexity of animation graphs that in AAA titles like 'Assassin's Creed' can handle over 15,000 animations and 12 nested graph layers. The framework converts user speed, direction, style specifications, and in-game events into keyframes, and the neural backbone progressively generates timing, root trajectory, pose tokens, and final continuous motion. On the high-level side, 'smart locomotion' handles movement and 'smart object' handles interactions with objects and scenes, generating styles such as limping, zombie-like movement, skipping, and strafing in a zero-shot manner from a single prompt.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.