List of all items

Structs

• Context
• Device
• DeviceBuffer
• DeviceSlice
• Dim3
• Event
• Function
• JitDiagnostic
• JitLog
• JitOptions
• Kernel
• LaunchParams
• LaunchParamsBuilder
• Module
• PinnedBuffer
• Stream
• UnifiedBuffer
• collective::CollectiveConfig
• collective::CollectiveOps
• collective::CommGroup
• collective::Communicator
• collective::RingAllReduce
• collective::TreeAllReduce
• device_pool::DeviceStatus
• device_pool::GpuLease
• device_pool::GpuTask
• device_pool::MultiGpuPool
• device_pool::WorkloadBalancer
• distributed::Bucket
• distributed::DataParallelConfig
• distributed::DistributedConfig
• distributed::DistributedOptimizer
• distributed::DistributedRuntime
• distributed::FileStore
• distributed::GradientBucket
• distributed::ModelParallelConfig
• distributed::NodeId
• distributed::NodeInfo
• distributed::ProcessGroup
• distributed::TcpStore
• global_init::OxiCudaRuntime
• global_init::OxiCudaRuntimeBuilder
• global_init::OxiCudaRuntimeConfig
• global_init::RuntimeInfo
• pipeline_parallel::ActivationCheckpointing
• pipeline_parallel::BubbleAnalysis
• pipeline_parallel::GpipeScheduler
• pipeline_parallel::InterleavedScheduler
• pipeline_parallel::OneFOneBScheduler
• pipeline_parallel::PipelineConfig
• pipeline_parallel::PipelineEngine
• pipeline_parallel::PipelineEvent
• pipeline_parallel::PipelineStage
• pipeline_parallel::PipelineVisualizer
• pipeline_parallel::ZeroBubbleScheduler
• prelude::Context
• prelude::Device
• prelude::DeviceBuffer
• prelude::Dim3
• prelude::Event
• prelude::Function
• prelude::Kernel
• prelude::LaunchParams
• prelude::Module
• prelude::PinnedBuffer
• prelude::Stream
• prelude::UnifiedBuffer
• profiling::ChromeTraceExporter
• profiling::GlobalProfiler
• profiling::KernelProfiler
• profiling::ProfileSummary
• profiling::ProfilingSession
• profiling::SpanGuard
• profiling::TraceEvent
• profiling::TraceRecorder

Enums

• CudaError
• DriverLoadError
• JitSeverity
• collective::AllReduceAlgorithm
• collective::DataType
• collective::ReduceOp
• device_pool::DeviceSelectionPolicy
• distributed::DistributedBackend
• distributed::DistributedStatus
• distributed::InitMethod
• global_init::DeviceSelection
• pipeline_parallel::CheckpointDecision
• pipeline_parallel::EventType
• pipeline_parallel::MicrobatchStatus
• pipeline_parallel::PipelineSchedule
• prelude::CudaError
• profiling::TracePhase

Traits

• KernelArgs
• prelude::KernelArgs

Macros

• launch

Functions

• best_device
• copy::copy_dtod
• copy::copy_dtod_async
• copy::copy_dtoh
• copy::copy_dtoh_async
• copy::copy_dtoh_async_raw
• copy::copy_htod
• copy::copy_htod_async
• copy::copy_htod_async_raw
• global_init::default_context
• global_init::default_device
• global_init::default_stream
• global_init::init_with
• global_init::is_initialized
• global_init::lazy_init
• global_init::runtime_info
• grid_size_for
• init
• list_devices
• prelude::best_device
• prelude::grid_size_for
• prelude::list_devices
• prelude::try_driver
• try_driver

Type Aliases

• CudaResult
• prelude::CudaResult

Constants

• AUTO_SELECT_THRESHOLD_BYTES
• SUPPORTED_ONNX_OPS
• features::HAS_AUTOTUNE
• features::HAS_BACKEND
• features::HAS_BLAS
• features::HAS_DNN
• features::HAS_FFT
• features::HAS_GLOBAL_INIT
• features::HAS_GPU_TESTS
• features::HAS_LEVEL_ZERO
• features::HAS_METAL
• features::HAS_ONNX_BACKEND
• features::HAS_POOL
• features::HAS_PTX
• features::HAS_RAND
• features::HAS_ROCM
• features::HAS_SOLVER
• features::HAS_SPARSE
• features::HAS_TENSOR_BACKEND
• features::HAS_TRANSFORMER_BACKEND
• features::HAS_VULKAN
• features::HAS_WASM_BACKEND
• features::HAS_WEBGPU