Table of Contents

Namespace DotCompute.Backends.CUDA.Advanced

Classes

AdaOptimizations

RTX 2000 Ada Lovelace generation specific optimizations and utilities

BottleneckAnalysis

Bottleneck analysis results

ConvolutionParams

Convolution parameters.

CudaCooperativeGroupsAnalysis

A class that represents cuda cooperative groups analysis.

CudaCooperativeGroupsManager

Manager for CUDA Cooperative Groups functionality

CudaCooperativeKernel

A class that represents cuda cooperative kernel.

CudaCooperativeLaunchConfig

A class that represents cuda cooperative launch config.

CudaCooperativeLaunchResult

A class that represents cuda cooperative launch result.

CudaDynamicParallelismManager

Manager for CUDA Dynamic Parallelism functionality

CudaKernelProfiler

Advanced kernel profiler for CUDA with RTX 2000 Ada optimizations

CudaTensorCoreAnalysis

A class that represents cuda tensor core analysis.

CudaTensorCoreExecutionMetrics

A class that represents cuda tensor core execution metrics.

CudaTensorCoreExecutionResult

A class that represents cuda tensor core execution result.

CudaTensorCoreKernel

A class that represents cuda tensor core kernel.

CudaTensorCoreManager

Manager for CUDA Tensor Core operations (RTX 2000 Ada specific)

CudaTensorCoreManagerProduction

Production-grade CUDA Tensor Core manager with WMMA operations, mixed precision support, and performance profiling.

CudaTensorDescriptor

A class that represents cuda tensor descriptor.

CudaTensorGEMMOperation

A class that represents cuda tensor g e m m operation.

CudaTensorMemoryLayout

A class that represents cuda tensor memory layout.

CudaTensorOperation

A class that represents cuda tensor operation.

CudaThroughputMetrics

CUDA-specific throughput performance metrics

GridConfig

Grid configuration for optimal execution

OccupancyMetrics

Occupancy metrics for kernel execution

ProfilingStatistics

Profiling statistics container

SharedMemoryConfig

Shared memory configuration for Ada generation

TensorCoreCapabilities

Tensor core capabilities.

TensorCoreException

Exception for tensor core operations.

TensorCoreResult

Tensor core operation result.

TensorCoreStatistics

Tensor core statistics.

UnifiedValidationResult

Validation result for Ada configurations

Structs

WmmaShape

WMMA shape configuration.

dim3

CUDA dimension structure.

Enums

CudaCooperativeOptimizationLevel

An cuda cooperative optimization level enumeration.

CudaTensorFormat

An cuda tensor format enumeration.

CudaTensorOperationType

An cuda tensor operation type enumeration.

CudaTensorPrecision

An cuda tensor precision enumeration.

DataType

An data type enumeration.

MatrixLayout

An matrix layout enumeration.

SharedMemoryCarveout

An shared memory carveout enumeration.

WorkloadType

An workload type enumeration.