org.apfloat.internal

Class LongParallelFNTStrategy

java.lang.Object

org.apfloat.internal.LongElementaryModMath

org.apfloat.internal.LongModMath

org.apfloat.internal.LongTableFNTStrategy

org.apfloat.internal.LongParallelFNTStrategy

All Implemented Interfaces:

ParallelNTTStrategy, NTTStrategy

Direct Known Subclasses:

LongFactor3SixStepNTTStrategy, LongSixStepFNTStrategy, LongTwoPassFNTStrategy

public abstract class LongParallelFNTStrategy
extends LongTableFNTStrategy
implements ParallelNTTStrategy

Common methods to calculate Fast Number Theoretic Transforms in parallel using multiple threads.

Note that to get any performance gain from running many threads in parallel, the JVM must be executing native threads. If the JVM is running in green threads mode, there is no advantage of having multiple threads, as the JVM will in fact execute just one thread and divide its time to multiple simulated threads.

Version:

1.5.1

Author:

Mikko Tommila

Field Summary

Fields

Modifier and Type	Field and Description
protected ParallelRunner	parallelRunner The parallel runner.

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	LongParallelFNTStrategy() Default constructor.

Method Summary

Methods

Modifier and Type	Method and Description
protected void	multiplyElements(ArrayAccess arrayAccess, int startRow, int rows, int columns, long w, long scaleFactor) Multiply each matrix element (i, j) by wi * j * scaleFactor.
void	setParallelRunner(ParallelRunner parallelRunner) Set the parallel runner to be used when executing the transform.
protected void	transformRows(int length, int count, boolean isInverse, ArrayAccess arrayAccess, long[] wTable, int[] permutationTable) Transform the rows of the data matrix.

Methods inherited from class org.apfloat.internal.LongTableFNTStrategy

getTransformLength, inverseTableFNT, inverseTransform, tableFNT, transform

Methods inherited from class org.apfloat.internal.LongModMath

createWTable, getForwardNthRoot, getInverseNthRoot, modDivide, modInverse, modPow, negate

Methods inherited from class org.apfloat.internal.LongElementaryModMath

getModulus, modAdd, modMultiply, modSubtract, setModulus

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apfloat.spi.NTTStrategy

getTransformLength, inverseTransform, transform

Field Detail

parallelRunner

protected ParallelRunner parallelRunner

The parallel runner.

Constructor Detail

LongParallelFNTStrategy

protected LongParallelFNTStrategy()

Default constructor.

Method Detail

setParallelRunner

public void setParallelRunner(ParallelRunner parallelRunner)

Description copied from interface: ParallelNTTStrategy

Set the parallel runner to be used when executing the transform.

Specified by:

setParallelRunner in interface ParallelNTTStrategy

Parameters:

parallelRunner - The parallel runner.

multiplyElements

protected void multiplyElements(ArrayAccess arrayAccess,
                    int startRow,
                    int rows,
                    int columns,
                    long w,
                    long scaleFactor)
                         throws ApfloatRuntimeException

Multiply each matrix element (i, j) by wi * j * scaleFactor. The matrix size is n₁ x n₂.

Parameters:

arrayAccess - The memory array to multiply.

startRow - Which row in the whole matrix the starting row in the arrayAccess is.

rows - The number of rows in the arrayAccess to multiply.

columns - The number of columns in the matrix (= n₂).

w - The n:th root of unity (where n = n₁ * n₂).

scaleFactor - An extra factor by which all elements are multiplied.

Throws:

ApfloatRuntimeException

transformRows

protected void transformRows(int length,
                 int count,
                 boolean isInverse,
                 ArrayAccess arrayAccess,
                 long[] wTable,
                 int[] permutationTable)
                      throws ApfloatRuntimeException

Transform the rows of the data matrix. If only one processor is available, it runs all transforms in the current thread. If more than one processor are available, it dispatches the calculations to multiple threads to parallelize the calculation. The number of processors is determined using ApfloatContext.getNumberOfProcessors().

Parameters:

length - Length of one transform (one row).

count - Number of rows.

isInverse - true if an inverse transform is performed, false if a forward transform is performed.

arrayAccess - The memory array to split to rows and to transform.

wTable - Table of powers of n:th root of unity (where n is the transform length).

permutationTable - Table of permutation indexes, or null if no permutation should be done.

Throws:

ApfloatRuntimeException