cuda-0.8.0.0: FFI binding to the CUDA interface for programming NVIDIA GPUs

Copyright[2009..2017] Trevor L. McDonell
LicenseBSD
Safe HaskellNone
LanguageHaskell98

Foreign.CUDA.Driver.Device

Contents

Description

Device management for low-level driver interface

Synopsis

Device Management

newtype Device Source #

A CUDA device

Constructors

Device 

Fields

Instances

data DeviceProperties Source #

The properties of a compute device

Constructors

DeviceProperties 

Fields

data DeviceAttribute Source #

Device attributes

Constructors

MaxThreadsPerBlock 
MaxBlockDimX 
MaxBlockDimY 
MaxBlockDimZ 
MaxGridDimX 
MaxGridDimY 
MaxGridDimZ 
MaxSharedMemoryPerBlock 
SharedMemoryPerBlock 
TotalConstantMemory 
WarpSize 
MaxPitch 
MaxRegistersPerBlock 
RegistersPerBlock 
ClockRate 
TextureAlignment 
GpuOverlap 
MultiprocessorCount 
KernelExecTimeout 
Integrated 
CanMapHostMemory 
ComputeMode 
MaximumTexture1dWidth 
MaximumTexture2dWidth 
MaximumTexture2dHeight 
MaximumTexture3dWidth 
MaximumTexture3dHeight 
MaximumTexture3dDepth 
MaximumTexture2dLayeredWidth 
MaximumTexture2dArrayWidth 
MaximumTexture2dLayeredHeight 
MaximumTexture2dArrayHeight 
MaximumTexture2dLayeredLayers 
MaximumTexture2dArrayNumslices 
SurfaceAlignment 
ConcurrentKernels 
EccEnabled 
PciBusId 
PciDeviceId 
TccDriver 
MemoryClockRate 
GlobalMemoryBusWidth 
L2CacheSize 
MaxThreadsPerMultiprocessor 
AsyncEngineCount 
UnifiedAddressing 
MaximumTexture1dLayeredWidth 
MaximumTexture1dLayeredLayers 
CanTex2dGather 
MaximumTexture2dGatherWidth 
MaximumTexture2dGatherHeight 
MaximumTexture3dWidthAlternate 
MaximumTexture3dHeightAlternate 
MaximumTexture3dDepthAlternate 
PciDomainId 
TexturePitchAlignment 
MaximumTexturecubemapWidth 
MaximumTexturecubemapLayeredWidth 
MaximumTexturecubemapLayeredLayers 
MaximumSurface1dWidth 
MaximumSurface2dWidth 
MaximumSurface2dHeight 
MaximumSurface3dWidth 
MaximumSurface3dHeight 
MaximumSurface3dDepth 
MaximumSurface1dLayeredWidth 
MaximumSurface1dLayeredLayers 
MaximumSurface2dLayeredWidth 
MaximumSurface2dLayeredHeight 
MaximumSurface2dLayeredLayers 
MaximumSurfacecubemapWidth 
MaximumSurfacecubemapLayeredWidth 
MaximumSurfacecubemapLayeredLayers 
MaximumTexture1dLinearWidth 
MaximumTexture2dLinearWidth 
MaximumTexture2dLinearHeight 
MaximumTexture2dLinearPitch 
MaximumTexture2dMipmappedWidth 
MaximumTexture2dMipmappedHeight 
ComputeCapabilityMajor 
ComputeCapabilityMinor 
MaximumTexture1dMipmappedWidth 
StreamPrioritiesSupported 
GlobalL1CacheSupported 
LocalL1CacheSupported 
MaxSharedMemoryPerMultiprocessor 
MaxRegistersPerMultiprocessor 
ManagedMemory 
MultiGpuBoard 
MultiGpuBoardGroupId 
HostNativeAtomicSupported 
SingleToDoublePrecisionPerfRatio 
PageableMemoryAccess 
ConcurrentManagedAccess 
ComputePreemptionSupported 
CanUseHostPointerForRegisteredMem 
CU_DEVICE_ATTRIBUTE_MAX 

data Compute Source #

GPU compute capability, major and minor revision number respectively.

Constructors

Compute !Int !Int 

data InitFlag Source #

Possible option flags for CUDA initialisation. Dummy instance until the API exports actual option values.

capability :: Device -> IO Compute Source #

Return the compute compatibility revision supported by the device

props :: Device -> IO DeviceProperties Source #

Return the properties of the selected device