OpenUSD (Universal Scene Description) is a high-performance, open-source 3D scene description framework originally developed by Pixar, used to non-destructively compose, describe, and exchange complex scene data across diverse software applications and rendering pipelines.
OpenUSD supports .usda (ASCII), .usdc (binary), and .usdz (zipped) formats. Its layered composition model enables non-destructive editing across teams. Governed by the Alliance for OpenUSD (AOUSD), it is the standard for industrial simulation, digital twins, and AR distribution on Apple platforms.
OpenUSD's core strength lies in its ability to represent a 3D scene as structured, layered data rather than a single monolithic file. The fundamental components include Layered Composition — a scene composed of multiple layers, each describing a distinct aspect (geometry, animation, lighting) — and Primitives (Prims), the foundational building blocks that hold all scene data.
USD's Composition Arcs (references, variants, inherits) govern how layers are stacked and merged into the final scene, following non-destructive rules. This allows artists and tools to contribute simultaneously without overwriting base data. Schemas define standard elements like UsdPreviewSurface for materials, ensuring reliable transfer between compliant applications.
Originally developed by Pixar for feature film production, OpenUSD has grown significantly since its 2016 open sourcing. Key areas include industrial simulation and digital twins (NVIDIA Omniverse uses USD as its foundational data model), augmented reality (Apple adopted USDZ as the primary AR container for iOS and visionOS), and content authoring (Blender, Maya, Houdini, and Substance 3D all support native USD export).
OpenUSD is highly compatible with, yet distinct from, web-centric delivery formats. The USDZ variant packages a USD scene and its textures into a single optimized archive, ideal for mobile AR distribution. In relation to glTF: OpenUSD is typically used earlier in production for complex data composition and exchange, while glTF serves as the optimized, runtime-ready delivery format. The USD layer structure enables 3D streaming by allowing a base scene to stream quickly, followed by progressive enhancement layers.
GL transmission format (glTF) — The open standard for real-time 3D delivery on the web, complementary to USD in production-to-delivery pipelines.
Asset optimization — Transforming production assets into formats suitable for real-time rendering and web delivery.
3D streaming — Progressive network delivery of interactive 3D content, well-suited to USD's layered scene structure.
Streaming video reshaped media consumption; we're doing the same for 3D. Join a small team solving tough spatial streaming, content delivery, and developer experience challenges.
Technical deep-dives on adaptive spatial streaming, the infrastructure challenges of 3D at scale, and what we're learning as we build. Written by the team doing the work.
