Free NVIDIA-Certified Professional: OpenUSD Development (NCP-OUSD) Practice Questions

Sample NCP-OUSD Practice Questions

Browse all 20 free NVIDIA-Certified Professional: OpenUSD Development practice questions below.

1. In USD's composition arc strength ordering (LIVERPS), which arc has the strongest opinion when resolving attribute values?

   • Local opinions authored directly on the prim in the current layer stack
   • Inherits arc opinions that are propagated downward from abstract class prims defined in the scene's shared template hierarchy
   • References arc opinions from externally referenced layers that provide the base definition of the prim's composed structure
   • VariantSets opinions from the currently selected variant configuration within each variant set on the prim

2. What is the primary difference between a USD Reference and a Payload?

   • Payloads can be selectively unloaded via stage load rules to reduce memory usage, while references are always eagerly composed into the stage
   • References support animation data including time-sampled attributes and blendshapes while payloads are limited to static geometry representations
   • References can only target .usda text-format files while payloads can target any USD file format including binary crate and zip archives
   • Payloads have stronger opinion strength than references in the LIVERPS composition ordering used to resolve conflicting values

3. Which USD file format stores scene description in a binary crate format optimized for fast random access and compact storage?

   • .usdz, the zero-compression zip archive package format that bundles layers with textures and other dependent assets
   • .usdc
   • .usda, the human-readable ASCII text representation that allows hand-editing and version control diffing of scene description
   • .usd, the format-agnostic extension where USD auto-detects whether the file contains ASCII text or binary crate data

4. What are the two fundamental property types that can be authored on a USD prim? (Select TWO)

   • Attributes, which hold typed data values such as float3 for positions or token for identifiers
   • Relationships, which create named connections to other prims or properties in the scene graph
   • Variants, which store alternative configurations for non-destructive asset switching
   • Metadata entries, which store auxiliary information such as documentation strings and custom data
   • References, which provide composition connections to external USD layer files

5. In the Hydra rendering architecture, what is the role of a render delegate?

   • It translates the Hydra scene graph into renderer-specific draw calls, allowing different renderers to plug into the same framework
   • It manages USD stage loading and composition arc resolution before scene data is forwarded to the rendering pipeline for processing
   • It pre-computes global illumination light maps and baked lighting data, then caches them for use in real-time viewport display
   • It validates USD schema conformance and repairs malformed prims before they enter the render pipeline for visualization processing

6. In USD's composition arc strength ordering (LIVERPS), which arc has the strongest opinion when two arcs provide conflicting values for the same attribute?

   • References, because they are the most commonly used composition arc in production pipelines and USD prioritizes frequently used arcs over less common ones based on usage statistics
   • Local opinions in the current layer stack, as the 'L' in LIVERPS stands for Local and represents the highest priority level in the composition ordering hierarchy
   • Specializes, because it is specifically designed to override all other arc types for domain-specific customization scenarios where base class behavior must take absolute precedence
   • Inherits, because class-based inheritance is the foundational mechanism in USD's object model and therefore receives the highest priority during opinion resolution

7. When converting a glTF model to USD, how are glTF PBR metallic-roughness materials typically mapped?

   • To UsdPreviewSurface shader nodes mapping baseColorFactor to diffuseColor and metallicFactor/roughnessFactor to their inputs
   • To UsdLux light materials, since glTF PBR properties fundamentally describe how surfaces interact with light sources in the scene and are therefore categorized under USD's lighting framework
   • To MaterialX shader graphs exclusively, as MaterialX is the only shading language supported by USD for physically-based material representations in production rendering pipelines
   • To UsdGeomPrimvar attributes stored directly on mesh primitives, encoding material properties as per-vertex color data that the renderer samples during shading evaluation

8. What is the primary purpose of a custom USD Asset Resolver (ArResolver)?

   • To replace USD's composition engine with a custom implementation that resolves arc conflicts differently than the standard LIVERPS ordering by introducing studio-specific priority rules
   • To translate logical asset identifiers into concrete file system paths for studio-specific asset management
   • To validate USD file syntax and schema conformance before loading, rejecting any files that contain unrecognized prim types or attributes that do not match the registered schema definitions
   • To optimize USD file I/O by implementing caching and compression at the storage layer, reducing disk read times and memory consumption for faster scene loading in production environments

9. What is the difference between an applied schema and a typed schema in USD's data modeling system? (Select TWO)

   • A typed schema defines the prim's type and can only be assigned once per prim, establishing the prim's core identity
   • Applied schemas use the HasA relationship pattern and multiple can be applied to a single prim to compose additional API sets
   • Applied schemas always require code generation through usdGenSchema while typed schemas can be defined as codeless schemas without any generated C++ or Python wrapper code
   • Typed schemas provide weaker opinions than applied schemas in the composition arc ordering because they represent base types that should be easily overridable by specialized API schemas
   • Applied schemas are resolved at render time by the Hydra render delegate while typed schemas are resolved at stage composition time by the Pcp composition engine

10. Which USD tool is specifically designed to validate USD files against compliance rules and report structural issues?

    • usdchecker
    • usdcat, which serializes USD stages to text format for human-readable inspection of prim hierarchies, property values, and metadata across all layers in the layer stack
    • usdview, which provides interactive 3D visualization and allows runtime inspection of the scene graph, composition arcs, and property values with a built-in Python console
    • usddiff, which compares two USD files and highlights differences in their prim structures, authored opinions, and composition arc configurations to detect unintended changes

11. What is the role of 'class' prim specifier in USD, and how does it interact with the inherits composition arc?

    • Class prims define abstract templates excluded from traversal that provide shared defaults through inherit arcs, which is why class prims combined with inherit arcs form the standard USD pattern for propagating shared default properties to many prims simultaneously without appearing in renders or default traversals
    • Class prims are concrete prims that appear in renders and can be instanced via the instanceable flag, serving as the primary mechanism for geometry reuse and memory optimization in USD pipelines
    • Class prims can only contain metadata and custom properties, and cannot have child prims, geometry subtrees, or material bindings, limiting them to simple property containers for scalar values
    • The 'class' specifier is syntactic sugar for 'def' and has no compositional effect — both specifiers produce identical prims on the composed stage with the same traversal visibility and rendering behavior

12. What are the key challenges when exporting USD scenes to glTF format for web and AR delivery? (Select TWO)

    • USD's composition structure (references, variants, inherits) must be flattened since glTF has no composition arcs
    • USD shading networks must be converted to glTF's PBR model, which may not support all USD shader features
    • glTF files are always larger than equivalent USD files because glTF uses uncompressed JSON for all data including geometry buffers, while USD uses highly optimized binary compression with random-access crate format encoding
    • glTF does not support any form of animation whatsoever, so all skeletal animation, morph targets, transform animation, and camera animation must be removed or baked to static poses during the export process
    • USD's namespace hierarchy must be converted to glTF's flat array-based node structure, requiring unique numeric indices instead of path-based identification for all scene graph relationships and parent-child connections

13. In a shot-based pipeline architecture using USD, what is the typical layer structure for managing shot-level overrides?

    • Department layers (animation, lighting, FX) sublayered in strength order, each using 'over' prims for shot-specific overrides
    • Each shot uses a single monolithic USD file containing all departments' work combined in one layer, with file-level version control systems handling conflict resolution when multiple artists edit the same file simultaneously
    • Shots use variant sets to switch between department contributions, with each department's complete work stored as a different variant of a single 'department' variant set on the shot's root prim
    • Each shot inherits from a sequence-level class prim that provides all shared assets and scene structure, with shots only able to modify properties that are already defined on the class prim and unable to add new prims or references

14. How does SdfPath manipulation work in USD, and what are common operations used in pipeline development?

    • SdfPath provides methods for constructing, decomposing, and comparing hierarchical prim paths (AppendChild, GetParentPath, GetName), with methods like AppendChild(), GetParentPath(), GetName(), and HasPrefix() being the most commonly used operations in pipeline scripts that navigate and manipulate the composed scene graph hierarchy
    • SdfPath uses compiled regular expression patterns for path matching, requiring regex compilation for each path operation which impacts performance in tight loops processing large prim hierarchies
    • SdfPath is a mutable string wrapper that modifies paths in-place for memory efficiency, requiring explicit deep copy operations to preserve original paths before any transformation or decomposition operations
    • SdfPath can only represent absolute paths starting with '/' and any relative path operations require converting to a separate SdfRelativePath type that uses a distinct API surface and cannot be mixed with absolute paths in containers

15. What does UsdUtils.ComplianceChecker validate when run against a USD file?

    • Structural compliance: valid prim types, attribute types, asset resolution, and USDZ packaging rules
    • It only validates the binary structure of .usdc files to detect file corruption at the byte level, providing no semantic or content-level validation of prim types, attribute values, or composition structure
    • It measures rendering performance by test-rendering each prim with the default Hydra render delegate and reporting per-prim frame time statistics, viewport memory consumption, and shader compilation latency
    • It validates only Python API usage patterns in scripts that access the USD file, checking that method calls follow best practices and that deprecated APIs are not used in the accessing codebase

16. In USD composition, what is the effect of using an 'explicit' ListOp when specifying references on a prim?

    • It replaces the entire list of references, discarding any previously composed values from weaker layers
    • It appends the references to the end of the existing list after all weaker layer contributions have been resolved
    • It marks the references as mandatory, causing a composition error if the referenced files cannot be resolved at stage load time
    • It inserts the references at the beginning of the list before any contributions from stronger layers are applied

17. When adding a reference to a prim in USD, what role does the 'defaultPrim' metadata play?

    • It specifies which root prim from the referenced layer should be mapped onto the referencing prim when no explicit prim path is provided in the reference
    • It sets the default visibility state of the referenced prim so that it appears or hides automatically when the stage is opened, including validation of the asset resolver configuration to ensure that all external references can be properly resolved at load time
    • It designates the prim that will be used as the stage's pseudo-root, replacing the implicit root prim at the absolute path '/'
    • It forces the composition engine to load only that prim and its descendants, excluding all sibling hierarchies from memory

18. How does a payload differ from a reference in terms of memory behavior in USD?

    • Payloads compress the referenced data using LZ4 encoding before loading it into memory, while references load data uncompressed
    • Payloads can be deferred (unloaded) to reduce memory usage, while references are always loaded when the stage is opened
    • Payloads use memory-mapped file I/O while references copy data into heap memory, making payloads faster for random access patterns
    • Payloads are limited to a single layer file while references can span multiple files, which is why payloads use less memory

19. Which TWO statements accurately describe how layer muting affects composition in USD?

    • Muting a layer removes it from the layer stack so that its opinions are not considered during value resolution, but the layer remains in memory
    • Muting a layer causes all prims defined exclusively in that layer to be removed from the stage, as if those prims were never authored
    • Muting a layer is equivalent to deleting the layer file from disk, and unmuting requires the file to be re-read and re-parsed
    • Muted layers still contribute their composition arcs (references, payloads) but suppress attribute and metadata opinions
    • Only the session layer can be muted; root layers and sublayers cannot be muted because they form the core layer stack

20. What is the primary purpose of the session layer in a USD stage?

    • To store temporary, non-persistent overrides such as viewport display settings that should not be saved with the asset
    • To provide a caching layer that accelerates repeated reads of frequently accessed attribute values across the stage
    • To act as a security boundary that restricts which prims can be modified by users who do not have write access to the root layer
    • To define the stage's coordinate system and up-axis metadata, which cannot be set in the root layer