| Package | Description |
|---|---|
| jcuda.driver |
Contains the classes related to the JCuda driver API.
|
| jcuda.runtime |
Contains the classes related to the JCuda runtime API.
|
| jcuda.utils | |
| org.nd4j.linalg.jcublas.context | |
| org.nd4j.linalg.jcublas.context.pool | |
| org.nd4j.linalg.jcublas.context.pool.factory |
| Modifier and Type | Field and Description |
|---|---|
static CUstream |
JCudaDriver.CU_STREAM_LEGACY
Stream handle that can be passed as a CUstream to use an implicit stream
with legacy synchronization behavior.
|
static CUstream |
JCudaDriver.CU_STREAM_PER_THREAD
Stream handle that can be passed as a CUstream to use an implicit stream
with per-thread synchronization behavior.
|
| Modifier and Type | Method and Description |
|---|---|
void |
CUstreamCallback.call(CUstream hStream,
int status,
Object userData)
The function that will be called
|
static int |
JCudaDriver.cuEventRecord(CUevent hEvent,
CUstream hStream)
Records an event.
|
static int |
JCudaDriver.cuGLMapBufferObjectAsync(CUdeviceptr dptr,
long[] size,
int buffer,
CUstream hStream)
Deprecated.
Deprecated as of CUDA 3.0
|
static int |
JCudaDriver.cuGLUnmapBufferObjectAsync(int buffer,
CUstream hStream)
Deprecated.
Deprecated as of CUDA 3.0
|
static int |
JCudaDriver.cuGraphicsMapResources(int count,
CUgraphicsResource[] resources,
CUstream hStream)
Map graphics resources for access by CUDA.
|
static int |
JCudaDriver.cuGraphicsUnmapResources(int count,
CUgraphicsResource[] resources,
CUstream hStream)
Unmap graphics resources.
|
static int |
JCudaDriver.cuLaunchGridAsync(CUfunction f,
int grid_width,
int grid_height,
CUstream hStream)
Deprecated.
Deprecated in CUDA
|
static int |
JCudaDriver.cuLaunchKernel(CUfunction f,
int gridDimX,
int gridDimY,
int gridDimZ,
int blockDimX,
int blockDimY,
int blockDimZ,
int sharedMemBytes,
CUstream hStream,
Pointer kernelParams,
Pointer extra)
Launches a CUDA function.
|
static int |
JCudaDriver.cuMemcpy2DAsync(CUDA_MEMCPY2D pCopy,
CUstream hStream)
Copies memory for 2D arrays.
|
static int |
JCudaDriver.cuMemcpy3DAsync(CUDA_MEMCPY3D pCopy,
CUstream hStream)
Copies memory for 3D arrays.
|
static int |
JCudaDriver.cuMemcpy3DPeerAsync(CUDA_MEMCPY3D_PEER pCopy,
CUstream hStream)
Copies memory between contexts asynchronously.
|
static int |
JCudaDriver.cuMemcpyAsync(CUdeviceptr dst,
CUdeviceptr src,
long ByteCount,
CUstream hStream)
Copies memory asynchronously.
|
static int |
JCudaDriver.cuMemcpyAtoHAsync(Pointer dstHost,
CUarray srcArray,
long srcIndex,
long ByteCount,
CUstream hStream)
Copies memory from Array to Host.
|
static int |
JCudaDriver.cuMemcpyDtoDAsync(CUdeviceptr dstDevice,
CUdeviceptr srcDevice,
long ByteCount,
CUstream hStream)
Copies memory from Device to Device.
|
static int |
JCudaDriver.cuMemcpyDtoHAsync(Pointer dstHost,
CUdeviceptr srcDevice,
long ByteCount,
CUstream hStream)
Copies memory from Device to Host.
|
static int |
JCudaDriver.cuMemcpyHtoAAsync(CUarray dstArray,
long dstIndex,
Pointer pSrc,
long ByteCount,
CUstream hStream)
Copies memory from Host to Array.
|
static int |
JCudaDriver.cuMemcpyHtoDAsync(CUdeviceptr dstDevice,
Pointer srcHost,
long ByteCount,
CUstream hStream)
Copies memory from Host to Device.
|
static int |
JCudaDriver.cuMemcpyPeerAsync(CUdeviceptr dstDevice,
CUcontext dstContext,
CUdeviceptr srcDevice,
CUcontext srcContext,
long ByteCount,
CUstream hStream)
Copies device memory between two contexts asynchronously.
|
static int |
JCudaDriver.cuMemsetD16Async(CUdeviceptr dstDevice,
short us,
long N,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuMemsetD2D16Async(CUdeviceptr dstDevice,
long dstPitch,
short us,
long Width,
long Height,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuMemsetD2D32Async(CUdeviceptr dstDevice,
long dstPitch,
int ui,
long Width,
long Height,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuMemsetD2D8Async(CUdeviceptr dstDevice,
long dstPitch,
byte uc,
long Width,
long Height,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuMemsetD32Async(CUdeviceptr dstDevice,
int ui,
long N,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuMemsetD8Async(CUdeviceptr dstDevice,
byte uc,
long N,
CUstream hStream)
Sets device memory.
|
static int |
JCudaDriver.cuStreamAddCallback(CUstream hStream,
CUstreamCallback callback,
Object userData,
int flags)
Add a callback to a compute stream.
|
static int |
JCudaDriver.cuStreamAttachMemAsync(CUstream hStream,
CUdeviceptr dptr,
long length,
int flags) |
static int |
JCudaDriver.cuStreamCreate(CUstream phStream,
int Flags)
Create a stream.
|
static int |
JCudaDriver.cuStreamCreateWithPriority(CUstream phStream,
int flags,
int priority) |
static int |
JCudaDriver.cuStreamDestroy(CUstream hStream)
Destroys a stream.
|
static int |
JCudaDriver.cuStreamGetFlags(CUstream hStream,
int[] flags) |
static int |
JCudaDriver.cuStreamGetPriority(CUstream hStream,
int[] priority) |
static int |
JCudaDriver.cuStreamQuery(CUstream hStream)
Determine status of a compute stream.
|
static int |
JCudaDriver.cuStreamSynchronize(CUstream hStream)
Wait until a stream's tasks are completed.
|
static int |
JCudaDriver.cuStreamWaitEvent(CUstream hStream,
CUevent hEvent,
int Flags)
Make a compute stream wait on an event.
|
| Constructor and Description |
|---|
cudaStream_t(CUstream stream)
Creates a cudaStream_t for the given
CUstream. |
| Modifier and Type | Method and Description |
|---|---|
KernelLauncher |
KernelLauncher.setStream(CUstream stream)
Set the stream for the function call.
This corresponds to the fourth parameter in the runtime call:
kernel<<<gridSize, blockSize,
sharedMemSize, stream>>>(...);
The default stream is null (0). |
KernelLauncher |
KernelLauncher.setup(dim3 gridSize,
dim3 blockSize,
int sharedMemSize,
CUstream stream)
Set the given grid size and block size, shared memory size
and stream for this KernelLauncher.
|
| Modifier and Type | Method and Description |
|---|---|
CUstream |
ContextHolder.getStream()
Get the stream for the current thread
based on the device for the thread
|
| Modifier and Type | Method and Description |
|---|---|
org.apache.commons.pool2.ObjectPool<CUstream> |
ContextHolder.getStreamPool() |
| Constructor and Description |
|---|
StreamPool(org.apache.commons.pool2.PooledObjectFactory<CUstream> factory) |
StreamPool(org.apache.commons.pool2.PooledObjectFactory<CUstream> factory,
org.apache.commons.pool2.impl.GenericObjectPoolConfig config) |
StreamPool(org.apache.commons.pool2.PooledObjectFactory<CUstream> factory,
org.apache.commons.pool2.impl.GenericObjectPoolConfig config,
org.apache.commons.pool2.impl.AbandonedConfig abandonedConfig) |
| Modifier and Type | Method and Description |
|---|---|
CUstream |
StreamItemFactory.create() |
| Modifier and Type | Method and Description |
|---|---|
org.apache.commons.pool2.PooledObject<CUstream> |
StreamItemFactory.wrap(CUstream cUstream) |
| Modifier and Type | Method and Description |
|---|---|
org.apache.commons.pool2.PooledObject<CUstream> |
StreamItemFactory.wrap(CUstream cUstream) |
| Modifier and Type | Method and Description |
|---|---|
void |
StreamItemFactory.destroyObject(org.apache.commons.pool2.PooledObject<CUstream> p) |
Copyright © 2015. All Rights Reserved.