org.apache.sysml.runtime.matrix.data

Class LibMatrixCuDNN

java.lang.Object

org.apache.sysml.runtime.matrix.data.LibMatrixCUDA

org.apache.sysml.runtime.matrix.data.LibMatrixCuDNN

public class LibMatrixCuDNN
extends LibMatrixCUDA

This class contains method that invoke CuDNN operations.

Field Summary

Fields

Modifier and Type	Field and Description
protected static int	CONVOLUTION_PREFERENCE

Fields inherited from class org.apache.sysml.runtime.matrix.data.LibMatrixCUDA

cudaSupportFunctions, CUDNN_DATA_TYPE, customKernelSuffix, maxNumElementsOfCuDNNTensor, sizeOfDataType

Constructor Summary

Constructors

Constructor and Description
LibMatrixCuDNN()

Method Summary

Modifier and Type	Method and Description
protected static void	checkStatus(int status) Convenience method for checking the status of CuDNN kernel.
static void	conv2d(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter, org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, double intermediateMemoryBudget) Performs a 2D convolution
static void	conv2dBackwardData(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter, org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout, org.apache.sysml.runtime.controlprogram.caching.MatrixObject output, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, double intermediateMemoryBudget) This method computes the backpropogation errors for previous layer of convolution operation
static void	conv2dBackwardFilter(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout, org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, double intermediateMemoryBudget) This method computes the backpropogation errors for filter of convolution operation
static void	conv2dBiasAdd(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, org.apache.sysml.runtime.controlprogram.caching.MatrixObject bias, org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter, org.apache.sysml.runtime.controlprogram.caching.MatrixObject output, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, double intermediateMemoryBudget) Does a 2D convolution followed by a bias_add
protected static jcuda.jcudnn.cudnnHandle	getCudnnHandle(GPUContext gCtx)
protected static jcuda.Pointer	getDensePointerForCuDNN(GPUContext gCtx, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, String instName) Convenience method to get jcudaDenseMatrixPtr.
static void	pooling(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, LibMatrixDNN.PoolingType poolingType, double intermediateMemoryBudget) performs maxpooling on GPU by exploiting cudnnPoolingForward(...)
static void	poolingBackward(GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject image, org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout, org.apache.sysml.runtime.controlprogram.caching.MatrixObject maxpoolOutput, org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock, int N, int C, int H, int W, int K, int R, int S, int pad_h, int pad_w, int stride_h, int stride_w, int P, int Q, LibMatrixDNN.PoolingType poolingType, double intermediateMemoryBudget) Performs maxpoolingBackward on GPU by exploiting cudnnPoolingBackward(...) This method computes the backpropogation errors for previous layer of maxpooling operation
static void	relu(org.apache.sysml.runtime.controlprogram.context.ExecutionContext ec, GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject in, String outputName) Performs the relu operation on the GPU.
static void	softmax(org.apache.sysml.runtime.controlprogram.context.ExecutionContext ec, GPUContext gCtx, String instName, org.apache.sysml.runtime.controlprogram.caching.MatrixObject in1, String outputName) Performs an "softmax" operation on a matrix on the GPU

Methods inherited from class org.apache.sysml.runtime.matrix.data.LibMatrixCUDA

abs, acos, asin, atan, axpy, biasAdd, biasMultiply, cbind, ceil, channelSums, cos, cosh, dataTypePointerTo, double2float, exp, float2double, floor, getCublasHandle, getCudaKernels, getCusparseHandle, getDenseMatrixOutputForGPUInstruction, getDensePointer, getNnz, getSparsePointer, isInSparseFormat, log, matmultTSMM, matrixMatrixArithmetic, matrixMatrixRelational, matrixScalarArithmetic, matrixScalarOp, matrixScalarRelational, one, rbind, reluBackward, resetFloatingPointPrecision, round, sigmoid, sign, sin, sinh, sliceDenseDense, sliceOperations, sliceSparseDense, solve, sqrt, tan, tanh, toInt, transpose, unaryAggregate, zero

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

CONVOLUTION_PREFERENCE

protected static int CONVOLUTION_PREFERENCE

Constructor Detail

LibMatrixCuDNN

public LibMatrixCuDNN()

Method Detail

getCudnnHandle

protected static jcuda.jcudnn.cudnnHandle getCudnnHandle(GPUContext gCtx)
                                                  throws DMLRuntimeException

Throws:

DMLRuntimeException

conv2dBiasAdd

public static void conv2dBiasAdd(GPUContext gCtx,
                                 String instName,
                                 org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                                 org.apache.sysml.runtime.controlprogram.caching.MatrixObject bias,
                                 org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter,
                                 org.apache.sysml.runtime.controlprogram.caching.MatrixObject output,
                                 int N,
                                 int C,
                                 int H,
                                 int W,
                                 int K,
                                 int R,
                                 int S,
                                 int pad_h,
                                 int pad_w,
                                 int stride_h,
                                 int stride_w,
                                 int P,
                                 int Q,
                                 double intermediateMemoryBudget)
                          throws DMLRuntimeException

Does a 2D convolution followed by a bias_add

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

image - input image matrix object

bias - bias matrix object

filter - filter matrix object

output - output matrix object

N - number of input images

C - number of channels

H - height of each image

W - width of each image

K - number of output "channels"

R - height of filter

S - width of filter

pad_h - padding height

pad_w - padding width

stride_h - stride height

stride_w - string width

P - output height

Q - output width

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if error

conv2d

public static void conv2d(GPUContext gCtx,
                          String instName,
                          org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                          org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter,
                          org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock,
                          int N,
                          int C,
                          int H,
                          int W,
                          int K,
                          int R,
                          int S,
                          int pad_h,
                          int pad_w,
                          int stride_h,
                          int stride_w,
                          int P,
                          int Q,
                          double intermediateMemoryBudget)
                   throws DMLRuntimeException

Performs a 2D convolution

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

image - input matrix object

filter - filter matrix object

outputBlock - output matrix object

N - number of input images

C - number of channels

H - height of each image

W - width of each image

K - number of output "channels"

R - height of filter

S - width of filter

pad_h - padding height

pad_w - padding width

stride_h - stride height

stride_w - string width

P - output height

Q - output width

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if error

softmax

public static void softmax(org.apache.sysml.runtime.controlprogram.context.ExecutionContext ec,
                           GPUContext gCtx,
                           String instName,
                           org.apache.sysml.runtime.controlprogram.caching.MatrixObject in1,
                           String outputName)
                    throws DMLRuntimeException

Performs an "softmax" operation on a matrix on the GPU

Parameters:

ec - execution context

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

in1 - input matrix

outputName - output matrix name

Throws:

DMLRuntimeException - if DMLRuntimeException occurs

conv2dBackwardFilter

public static void conv2dBackwardFilter(GPUContext gCtx,
                                        String instName,
                                        org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                                        org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout,
                                        org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock,
                                        int N,
                                        int C,
                                        int H,
                                        int W,
                                        int K,
                                        int R,
                                        int S,
                                        int pad_h,
                                        int pad_w,
                                        int stride_h,
                                        int stride_w,
                                        int P,
                                        int Q,
                                        double intermediateMemoryBudget)
                                 throws DMLRuntimeException

This method computes the backpropogation errors for filter of convolution operation

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

image - input image

dout - errors from next layer

outputBlock - output errors

N - number of images

C - number of channels

H - height

W - width

K - number of filters

R - filter height

S - filter width

pad_h - pad height

pad_w - pad width

stride_h - stride height

stride_w - stride width

P - output activation height

Q - output activation width

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if DMLRuntimeException occurs

conv2dBackwardData

public static void conv2dBackwardData(GPUContext gCtx,
                                      String instName,
                                      org.apache.sysml.runtime.controlprogram.caching.MatrixObject filter,
                                      org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout,
                                      org.apache.sysml.runtime.controlprogram.caching.MatrixObject output,
                                      int N,
                                      int C,
                                      int H,
                                      int W,
                                      int K,
                                      int R,
                                      int S,
                                      int pad_h,
                                      int pad_w,
                                      int stride_h,
                                      int stride_w,
                                      int P,
                                      int Q,
                                      double intermediateMemoryBudget)
                               throws DMLRuntimeException

This method computes the backpropogation errors for previous layer of convolution operation

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

filter - filter used in conv2d

dout - errors from next layer

output - output errors

N - number of images

C - number of channels

H - height

W - width

K - number of filters

R - filter height

S - filter width

pad_h - pad height

pad_w - pad width

stride_h - stride height

stride_w - stride width

P - output activation height

Q - output activation width

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if DMLRuntimeException occurs

pooling

public static void pooling(GPUContext gCtx,
                           String instName,
                           org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                           org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock,
                           int N,
                           int C,
                           int H,
                           int W,
                           int K,
                           int R,
                           int S,
                           int pad_h,
                           int pad_w,
                           int stride_h,
                           int stride_w,
                           int P,
                           int Q,
                           LibMatrixDNN.PoolingType poolingType,
                           double intermediateMemoryBudget)
                    throws DMLRuntimeException

performs maxpooling on GPU by exploiting cudnnPoolingForward(...)

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

image - image as matrix object

outputBlock - output matrix

N - batch size

C - number of channels

H - height of image

W - width of image

K - number of filters

R - height of filter

S - width of filter

pad_h - vertical padding

pad_w - horizontal padding

stride_h - horizontal stride

stride_w - vertical stride

P - (H - R + 1 + 2*pad_h)/stride_h

Q - (W - S + 1 + 2*pad_w)/stride_w

poolingType - type of pooling

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if DMLRuntimeException occurs

poolingBackward

public static void poolingBackward(GPUContext gCtx,
                                   String instName,
                                   org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                                   org.apache.sysml.runtime.controlprogram.caching.MatrixObject dout,
                                   org.apache.sysml.runtime.controlprogram.caching.MatrixObject maxpoolOutput,
                                   org.apache.sysml.runtime.controlprogram.caching.MatrixObject outputBlock,
                                   int N,
                                   int C,
                                   int H,
                                   int W,
                                   int K,
                                   int R,
                                   int S,
                                   int pad_h,
                                   int pad_w,
                                   int stride_h,
                                   int stride_w,
                                   int P,
                                   int Q,
                                   LibMatrixDNN.PoolingType poolingType,
                                   double intermediateMemoryBudget)
                            throws DMLRuntimeException

Performs maxpoolingBackward on GPU by exploiting cudnnPoolingBackward(...) This method computes the backpropogation errors for previous layer of maxpooling operation

Parameters:

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

image - image as matrix object

dout - delta matrix, output of previous layer

maxpoolOutput - (optional and can be null) output of maxpool forward function

outputBlock - output matrix

N - batch size

C - number of channels

H - height of image

W - width of image

K - number of filters

R - height of filter

S - width of filter

pad_h - vertical padding

pad_w - horizontal padding

stride_h - horizontal stride

stride_w - vertical stride

P - (H - R + 1 + 2*pad_h)/stride_h

Q - (W - S + 1 + 2*pad_w)/stride_w

poolingType - type of pooling

intermediateMemoryBudget - intermediate memory budget

Throws:

DMLRuntimeException - if DMLRuntimeException occurs

relu

public static void relu(org.apache.sysml.runtime.controlprogram.context.ExecutionContext ec,
                        GPUContext gCtx,
                        String instName,
                        org.apache.sysml.runtime.controlprogram.caching.MatrixObject in,
                        String outputName)
                 throws DMLRuntimeException

Performs the relu operation on the GPU.

Parameters:

ec - currently active ExecutionContext

gCtx - a valid GPUContext

instName - the invoking instruction's name for record Statistics.

in - input matrix

outputName - name of the output matrix

Throws:

DMLRuntimeException - if an error occurs

getDensePointerForCuDNN

protected static jcuda.Pointer getDensePointerForCuDNN(GPUContext gCtx,
                                                       org.apache.sysml.runtime.controlprogram.caching.MatrixObject image,
                                                       String instName)
                                                throws DMLRuntimeException

Convenience method to get jcudaDenseMatrixPtr. This method explicitly converts sparse to dense format, so use it judiciously.

Parameters:

gCtx - a valid GPUContext

image - input matrix object

instName - name of the instruction

Returns:

jcuda pointer

Throws:

DMLRuntimeException - if error occurs while sparse to dense conversion

checkStatus

protected static void checkStatus(int status)
                           throws DMLRuntimeException

Convenience method for checking the status of CuDNN kernel.

Parameters:

status - status returned by CuDNN

Throws:

DMLRuntimeException - if status is not CUDNN_STATUS_SUCCESS