jcuda.runtime
Class cudaDeviceProp

java.lang.Object
  extended by jcuda.runtime.cudaDeviceProp

public class cudaDeviceProp
extends java.lang.Object

Java port of the cudaDeviceProp.

Most comments are taken from the CUDA reference manual.

See Also:
JCuda.cudaGetDeviceProperties(jcuda.runtime.cudaDeviceProp, int)

Field Summary
 int asyncEngineCount
          1 when the device can concurrently copy memory between host and device while executing a kernel.
 int canMapHostMemory
          Device can map host memory with cudaHostAlloc/cudaHostGetDevicePointer
 int clockRate
          The clock frequency in kilohertz;
 int computeMode
          Compute mode (See ::cudaComputeMode)
 int concurrentKernels
          Device can possibly execute multiple kernels concurrently
 int deviceOverlap
          Device can concurrently copy memory and execute a kernel.
 int ECCEnabled
          Device has ECC support enabled
 int integrated
          Device is integrated as opposed to discrete
 int kernelExecTimeoutEnabled
          Specified whether there is a run time limit on kernels
 int l2CacheSize
          L2 cache size in bytes
 int major
          Major revision number defining the device's compute capability;
 int[] maxGridSize
          The maximum sizes of each dimension of a grid;
 int maxSurface1D
          Maximum 1D surface size
 int[] maxSurface1DLayered
          Maximum 1D layered surface dimensions
 int[] maxSurface2D
          Maximum 2D surface dimensions
 int[] maxSurface2DLayered
          Maximum 2D layered surface dimensions
 int[] maxSurface3D
          Maximum 3D surface dimensions
 int maxSurfaceCubemap
          Maximum Cubemap surface dimensions
 int[] maxSurfaceCubemapLayered
          Maximum Cubemap layered surface dimensions
 int maxTexture1D
          Maximum 1D texture size
 int[] maxTexture1DLayered
          Maximum 1D layered texture dimensions
 int maxTexture1DLinear
          Maximum size for 1D textures bound to linear memory
 int maxTexture1DMipmap
          Maximum 1D mipmapped texture size
 int[] maxTexture2D
          Maximum 2D texture dimensions
 int[] maxTexture2DGather
          Maximum 2D texture dimensions if texture gather operations have to be performed
 int[] maxTexture2DLayered
          Maximum 2D layered texture dimensions
 int[] maxTexture2DLinear
          Maximum dimensions (width, height, pitch) for 2D textures bound to pitched memory
 int[] maxTexture2DMipmap
          Maximum 2D mipmapped texture dimensions
 int[] maxTexture3D
          Maximum 3D texture dimensions
 int maxTextureCubemap
          Maximum Cubemap texture dimensions
 int[] maxTextureCubemapLayered
          Maximum Cubemap layered texture dimensions
 int[] maxThreadsDim
          The maximum sizes of each dimension of a block;
 int maxThreadsPerBlock
          The maximum number of threads per block;
 int maxThreadsPerMultiProcessor
          The number of maximum resident threads per multiprocessor.
 int memoryBusWidth
          The memory bus width in bits
 int memoryClockRate
          The peak memory clock frequency in kilohertz.
 long memPitch
          The maximum pitch in bytes allowed by the memory copy functions that involve memory regions allocated through cudaMallocPitch();
 int minor
          Minor revision number defining the device's compute capability;
 int multiProcessorCount
          Number of multiprocessors on device
 byte[] name
          An ASCII string identifying the device;
 int pciBusID
          PCI bus ID of the device
 int pciDeviceID
          PCI device ID of the device
 int pciDomainID
          PCI domain ID of the device
 int regsPerBlock
          The maximum number of 32-bit registers available to a thread block; this number is shared by all thread blocks simultaneously resident on a multiprocessor;
 long sharedMemPerBlock
          The maximum amount of shared memory available to a thread block in bytes; this amount is shared by all thread blocks simultaneously resident on a multiprocessor;
 long surfaceAlignment
          Alignment requirements for surfaces
 int tccDriver
          1 if device is a Tesla device using TCC driver, 0 otherwise
 long textureAlignment
          The alignment requirement; texture base addresses that are aligned to textureAlignment bytes do not need an offset applied to texture fetches;
 long texturePitchAlignment
          Pitch alignment requirement for texture references bound to pitched memory
 long totalConstMem
          The total amount of constant memory available on the device in bytes;
 long totalGlobalMem
          The total amount of global memory available on the device in bytes;
 int unifiedAddressing
          1 if the device shares a unified address space with the host and 0 otherwise.
 int warpSize
          The warp size in threads;
 
Constructor Summary
cudaDeviceProp()
          Creates a new, uninitialized cudaDeviceProp object
 
Method Summary
 java.lang.String getName()
          Returns the String describing the name of this cudaDeviceProp
 void setName(java.lang.String nameString)
          Set the name of this cudaDeviceProp to the given name
 java.lang.String toFormattedString()
          Creates and returns a formatted (aligned, multi-line) String representation of this object
 java.lang.String toString()
          Returns a String representation of this object.
 
Methods inherited from class java.lang.Object
equals, getClass, hashCode, notify, notifyAll, wait, wait, wait
 

Field Detail

name

public byte[] name
An ASCII string identifying the device;


totalGlobalMem

public long totalGlobalMem
The total amount of global memory available on the device in bytes;


sharedMemPerBlock

public long sharedMemPerBlock
The maximum amount of shared memory available to a thread block in bytes; this amount is shared by all thread blocks simultaneously resident on a multiprocessor;


regsPerBlock

public int regsPerBlock
The maximum number of 32-bit registers available to a thread block; this number is shared by all thread blocks simultaneously resident on a multiprocessor;


warpSize

public int warpSize
The warp size in threads;


memPitch

public long memPitch
The maximum pitch in bytes allowed by the memory copy functions that involve memory regions allocated through cudaMallocPitch();


maxThreadsPerBlock

public int maxThreadsPerBlock
The maximum number of threads per block;


maxThreadsDim

public int[] maxThreadsDim
The maximum sizes of each dimension of a block;


maxGridSize

public int[] maxGridSize
The maximum sizes of each dimension of a grid;


clockRate

public int clockRate
The clock frequency in kilohertz;


totalConstMem

public long totalConstMem
The total amount of constant memory available on the device in bytes;


major

public int major
Major revision number defining the device's compute capability;


minor

public int minor
Minor revision number defining the device's compute capability;


textureAlignment

public long textureAlignment
The alignment requirement; texture base addresses that are aligned to textureAlignment bytes do not need an offset applied to texture fetches;


texturePitchAlignment

public long texturePitchAlignment
Pitch alignment requirement for texture references bound to pitched memory


deviceOverlap

public int deviceOverlap
Device can concurrently copy memory and execute a kernel. Deprecated. Use instead asyncEngineCount.


multiProcessorCount

public int multiProcessorCount
Number of multiprocessors on device


kernelExecTimeoutEnabled

public int kernelExecTimeoutEnabled
Specified whether there is a run time limit on kernels


integrated

public int integrated
Device is integrated as opposed to discrete


canMapHostMemory

public int canMapHostMemory
Device can map host memory with cudaHostAlloc/cudaHostGetDevicePointer


computeMode

public int computeMode
Compute mode (See ::cudaComputeMode)


maxTexture1D

public int maxTexture1D
Maximum 1D texture size


maxTexture1DMipmap

public int maxTexture1DMipmap
Maximum 1D mipmapped texture size


maxTexture1DLinear

public int maxTexture1DLinear
Maximum size for 1D textures bound to linear memory


maxTexture2D

public int[] maxTexture2D
Maximum 2D texture dimensions


maxTexture2DMipmap

public int[] maxTexture2DMipmap
Maximum 2D mipmapped texture dimensions


maxTexture2DLinear

public int[] maxTexture2DLinear
Maximum dimensions (width, height, pitch) for 2D textures bound to pitched memory


maxTexture2DGather

public int[] maxTexture2DGather
Maximum 2D texture dimensions if texture gather operations have to be performed


maxTexture3D

public int[] maxTexture3D
Maximum 3D texture dimensions


maxTextureCubemap

public int maxTextureCubemap
Maximum Cubemap texture dimensions


maxTexture1DLayered

public int[] maxTexture1DLayered
Maximum 1D layered texture dimensions


maxTexture2DLayered

public int[] maxTexture2DLayered
Maximum 2D layered texture dimensions


maxTextureCubemapLayered

public int[] maxTextureCubemapLayered
Maximum Cubemap layered texture dimensions


maxSurface1D

public int maxSurface1D
Maximum 1D surface size


maxSurface2D

public int[] maxSurface2D
Maximum 2D surface dimensions


maxSurface3D

public int[] maxSurface3D
Maximum 3D surface dimensions


maxSurface1DLayered

public int[] maxSurface1DLayered
Maximum 1D layered surface dimensions


maxSurface2DLayered

public int[] maxSurface2DLayered
Maximum 2D layered surface dimensions


maxSurfaceCubemap

public int maxSurfaceCubemap
Maximum Cubemap surface dimensions


maxSurfaceCubemapLayered

public int[] maxSurfaceCubemapLayered
Maximum Cubemap layered surface dimensions


surfaceAlignment

public long surfaceAlignment
Alignment requirements for surfaces


concurrentKernels

public int concurrentKernels
Device can possibly execute multiple kernels concurrently


ECCEnabled

public int ECCEnabled
Device has ECC support enabled


pciBusID

public int pciBusID
PCI bus ID of the device


pciDeviceID

public int pciDeviceID
PCI device ID of the device


pciDomainID

public int pciDomainID
PCI domain ID of the device


tccDriver

public int tccDriver
1 if device is a Tesla device using TCC driver, 0 otherwise


asyncEngineCount

public int asyncEngineCount
1 when the device can concurrently copy memory between host and device while executing a kernel. It is 2 when the device can concurrently copy memory between host and device in both directions and execute a kernel at the same time. It is 0 if neither of these is supported.


unifiedAddressing

public int unifiedAddressing
1 if the device shares a unified address space with the host and 0 otherwise.


memoryClockRate

public int memoryClockRate
The peak memory clock frequency in kilohertz.


memoryBusWidth

public int memoryBusWidth
The memory bus width in bits


l2CacheSize

public int l2CacheSize
L2 cache size in bytes


maxThreadsPerMultiProcessor

public int maxThreadsPerMultiProcessor
The number of maximum resident threads per multiprocessor.

Constructor Detail

cudaDeviceProp

public cudaDeviceProp()
Creates a new, uninitialized cudaDeviceProp object

Method Detail

getName

public java.lang.String getName()
Returns the String describing the name of this cudaDeviceProp

Returns:
String The String describing the name of this cudaDeviceProp

setName

public void setName(java.lang.String nameString)
Set the name of this cudaDeviceProp to the given name

Parameters:
nameString - The name for this cudaDeviceProp

toString

public java.lang.String toString()
Returns a String representation of this object.

Overrides:
toString in class java.lang.Object
Returns:
A String representation of this object.

toFormattedString

public java.lang.String toFormattedString()
Creates and returns a formatted (aligned, multi-line) String representation of this object

Returns:
A formatted String representation of this object