GPU Interop: Cuda Entry#

Entry node for post-processing hydra render results for a single view. It outputs the properties of a RenderVar as attributes accessible to downstream nodes. All downstream nodes are scheduled as CUDA commands.

Installation#

To use this node enable omni.graph.image.nodes in the Extension Manager.

Inputs#

Name	Type	Descripton	Default
Source Name (inputs:sourceName)	`string`	The name of the source RenderVar (AOV). The properties of this AOV are the output of this node.	ldrColor

Outputs#

Name	Type	Descripton	Default
Buffer Size (outputs:bufferSize)	`uint`	Size of the buffer, if the selected RenderVar is a buffer, otherwise it represents the texture depth.	None
RenderVar Data (outputs:cudaMipmappedArray)	`uint64`	The RenderVar data, which is represented as a CUDA Mipmapped Array.	None
External Time Of Simulation Frame (outputs:externalTimeOfSimFrame)	`int64`	The external time on the master node, matching the simulation frame used to render this frame. It is only valid if the simulation and rendering are performed at the same, or different but constant rates.	None
Texture Format (outputs:format)	`uint64`	The texture format of the input image, matching the values of the enum carb::graphics::Format. If the RenderVar is a buffer, the value is unknown.	None
Frame ID (outputs:frameId)	`int64`	The frame number. It is only valid if the simulation and rendering are performed at the same, or different but constant rates.	None
Height (outputs:height)	`uint`	The width of the selected texture RenderVar. If the RenderVar is an arbitrary buffer, it is 0.	None
Renderer Time (outputs:hydraTime)	`double`	The time difference since the last rendered frame.	None
Is Buffer (outputs:isBuffer)	`bool`	True if the entry exposes an arbitrary buffer RenderVar as opposed to a texture RenderVar.	None
Mip Count (outputs:mipCount)	`uint`	The number of levels of detail encoded in ‘RenderVar Data’. If the RenderVar is not a texture, the value is 0.	None
Rational Time Of Sim Denominator (outputs:rationalTimeOfSimDenominator)	`uint64`	The denominator of the simulation time expressed as rational time. It is always valid and may be used to fetch data from Fabric during the node compute.	None
Rational Time Of Sim Numerator (outputs:rationalTimeOfSimNumerator)	`int64`	The numerator of the simulation time expressed as rational time. It is always valid and may be used to fetch data from Fabric during the node compute.	None
Simulation Time (outputs:simTime)	`double`	The time difference since the last simulation frame. May suffer from loss of precision and cannot be used as a deterministic identifier, for example for fetching data from Fabric during the node compute - use the Rational Time instead in that case.	None
Stream (outputs:stream)	`uint64`	The CUDA Stream used to order the CUDA commands scheduled by this graph execution.	None
Width (outputs:width)	`uint`	The width of the selected texture RenderVar. If the RenderVar is an arbitrary buffer, it is 0.	None

Metadata#

Name	Value
Unique ID	omni.graph.nodes.GpuInteropCudaEntry
Version	3
Extension	omni.graph.image.nodes
Has State?	False
Implementation Language	C++
Default Memory Type	cpu
Generated Code Exclusions	None
uiName	GPU Interop: Cuda Entry
Categories	graph:postRender,rendering
Generated Class Name	OgnGpuInteropCudaEntryDatabase
Python Module	omni.graph.image.nodes