Namespace DotCompute.Backends.CUDA.Types
Classes
- CacheStatistics
Cache statistics for CUDA operations
- CachedKernel
Represents cached kernel compilation data.
- CudaGraphExecutionResult
Result of CUDA graph execution.
- CudaGraphOptimizationOptions
Options for CUDA graph optimization.
- CudaGraphStatistics
CUDA graph statistics.
- CudaGraphUpdateParameters
CUDA graph update parameters.
- CudaKernelArguments
Kernel arguments for CUDA execution
- CudaKernelFusionOptions
CUDA kernel fusion options.
- CudaMemoryAlignment
CUDA memory alignment requirements
- CudaMemoryBufferInfo
CUDA unified memory buffer implementation.
- CudaMemoryStatisticsExtended
Extended memory statistics for CUDA memory management
- CudaMemoryTracker
CUDA memory manager for device memory allocation and management.
- DeviceInfo
Represents CUDA device information.
- KernelCompilationException
Exception thrown during kernel compilation
- KernelFusionCandidate
Kernel fusion candidate.
- TensorCoreConfig
Tensor core configuration.
- UnifiedValidationResult
Validation result for CUDA operations
- ValidationOptions
Validation options for CUDA operations.
Structs
- CudaExtent
Represents the extent (dimensions) of a CUDA array or texture.
- CudaMemLocation
CUDA memory location.
- CudaMemPoolProps
Properties for CUDA memory pool configuration.
- CudaStream
Represents a CUDA stream for asynchronous operations.
Enums
- CacheConfig
An cache config enumeration.
- CudaArchitecture
Represents NVIDIA GPU architectures for optimization targeting. Each architecture has specific capabilities and optimization strategies.
- CudaGraphCaptureMode
Defines the capture modes for CUDA graph recording. Different modes provide varying levels of thread safety and flexibility.
- CudaGraphOptimizationLevel
Defines the optimization levels available for CUDA graph operations. Higher levels provide more aggressive optimizations but may increase compilation time.
- CudaKernelType
Categorizes CUDA kernels by their computational pattern. Used for optimization decisions and kernel fusion strategies.
- CudaMemAllocationType
CUDA memory allocation type.
- CudaMemHandleType
CUDA memory handle type.
- CudaMemLocationType
CUDA memory location type.
- CudaMemPoolAttribute
CUDA memory pool attribute.
- CudaMemoryFlags
An cuda memory flags enumeration.
- CudaMemoryType
An cuda memory type enumeration.
- DataType
An data type enumeration.
- ExportFormat
Export formats for validation reports.
- ManagedMemoryOptions
Flags for CUDA managed memory allocation and behavior.
- MemoryResidence
Defines where unified memory is currently resident.
- ValidationStatus
Status of a validation test.
- WarpSchedulingMode
An warp scheduling mode enumeration.