Maximum number of planes per frame

CUDA IPC Handle Size

Must be set in cudaExternalMemoryGetMappedMipmappedArray if the mipmapped array is used as a color target in a graphics API

Must be set in cudaMalloc3DArray to create a cubemap CUDA array

Default CUDA array allocation flag

Must be set in cudaMallocArray, cudaMalloc3DArray or cudaMallocMipmappedArray in order to create a deferred mapping CUDA array or CUDA mipmapped array

Must be set in cudaMalloc3DArray to create a layered CUDA array

Must be set in cudaMallocArray, cudaMalloc3DArray or cudaMallocMipmappedArray in order to create a sparse CUDA array or CUDA mipmapped array

Indicates that the layered sparse CUDA array or CUDA mipmapped array has a single mip tail region for all layers

Must be set in cudaMallocArray or cudaMalloc3DArray in order to bind surfaces to the CUDA array

Must be set in cudaMallocArray or cudaMalloc3DArray in order to perform texture gather operations on the CUDA array

If set, any subsequent work pushed in a stream that participated in a call to cudaLaunchCooperativeKernelMultiDevice will only wait for the kernel launched on the GPU corresponding to that stream to complete before it begins execution.

If set, each kernel launched as part of cudaLaunchCooperativeKernelMultiDevice only waits for prior work in the stream corresponding to that GPU to complete before the kernel begins execution.

Device id that represents the CPU

Device flag - Keep local memory allocation after launch

Device flag - Support mapped pinned allocations

Device flags mask

Device flag - Automatic scheduling

Device flag - Use blocking synchronization

Device schedule flags mask

Device flag - Spin default scheduling

Device flag - Yield default scheduling

Device flag - Ensure synchronous memory operations on this context will synchronize

Event uses blocking synchronization

Default event flag

Event will not record timing data

Event is suitable for interprocess use. cudaEventDisableTiming must be set

Default event record flag

Event is captured in the graph as an external event node when performing stream capture

Default event wait flag

Event is captured in the graph as an external event node when performing stream capture

Indicates that the external memory object is a dedicated resource

When the /p flags parameter of cudaExternalSemaphoreSignalParams contains this flag, it indicates that signaling an external semaphore object should skip performing appropriate memory synchronization operations over all the external memory objects that are imported as cudaExternalMemoryHandleTypeNvSciBuf, which otherwise are performed by default to ensure data coherency with other importers of the same NvSciBuf memory objects.

When the /p flags parameter of cudaExternalSemaphoreWaitParams contains this flag, it indicates that waiting an external semaphore object should skip performing appropriate memory synchronization operations over all the external memory objects that are imported as cudaExternalMemoryHandleTypeNvSciBuf, which otherwise are performed by default to ensure data coherency with other importers of the same NvSciBuf memory objects.

This port activates when the kernel has finished executing.

This port activates when all blocks of the kernel have begun execution. See also cudaLaunchAttributeLaunchCompletionEvent.

This port activates when all blocks of the kernel have performed cudaTriggerProgrammaticLaunchCompletion() or have terminated. It must be used with edge type cudaGraphDependencyTypeProgrammatic. See also cudaLaunchAttributeProgrammaticEvent.

Default page-locked allocation flag

Map allocation into device space

Pinned memory accessible by all CUDA contexts

Write-combined memory

Default host memory registration flag

Memory-mapped I/O space

Map registered memory into device space

Pinned memory accessible by all CUDA contexts

Memory-mapped read-only

Tell the CUDA runtime that DeviceFlags is being set in cudaInitDevice call

Device id that represents an invalid device

Automatically enable peer access between remote devices as needed

Memory can be accessed by any stream on any device

Memory cannot be accessed by any stream on any device

Memory can only be accessed by a single stream on the associated device

This flag, if set, indicates that the memory will be used as a buffer for hardware accelerated decompression.

When /p flags of cudaDeviceGetNvSciSyncAttributes is set to this, it indicates that application need signaler specific NvSciSyncAttr to be filled by cudaDeviceGetNvSciSyncAttributes.

When /p flags of cudaDeviceGetNvSciSyncAttributes is set to this, it indicates that application need waiter specific NvSciSyncAttr to be filled by cudaDeviceGetNvSciSyncAttributes.

Default behavior

Assume global caching is enabled and cannot be automatically turned off

Default peer addressing enable flag

Default stream flag

Legacy stream handle

Stream handle that can be passed as a cudaStream_t to use an implicit stream with legacy synchronization behavior.

See details of the synchronization behavior.

Stream does not synchronize with stream 0 (the NULL stream)

Per-thread stream handle

Stream handle that can be passed as a cudaStream_t to use an implicit stream with per-thread synchronization behavior.

See details of the synchronization behavior.