Package jdistlib.accelerator

Class CpuComputeBackend

java.lang.Object

jdistlib.accelerator.CpuComputeBackend

All Implemented Interfaces:

AutoCloseable, ComputeBackend, LinearAlgebraBackend, SinglePrecisionLinearAlgebraBackend

public final class CpuComputeBackend
extends Object
implements ComputeBackend

Deterministic CPU reference implementation for every accelerated primitive.

Constructor Summary

Constructors

Constructor	Description
CpuComputeBackend()

Method Summary

Modifier and Type	Method	Description
boolean	available()
double[]	axpy(double alpha, double[] x, double[] y)
ComputeCapabilities	capabilities()
void	close()
ComputeDeviceInfo	deviceInfo()	Returns stable backend, runtime, driver, and device provenance where available.
double	dot(double[] x, double[] y)
String	id()
LogisticRegressionBatchResult	logisticRegression(double[][] design, double[] outcomes, double[][] states, double priorPrecision)
double[][]	matrixMultiply(double[][] left, double[][] right)
PreparedLogisticRegression	prepareLogisticRegression(double[][] design, double[] outcomes)	Keeps reusable data in backend-optimal storage when the backend supports it.
PreparedTransposeProduct	prepareTransposeProduct(double[][] matrix)	Keeps a reusable row-by-feature matrix ready for repeated X'v batches.
double[]	unary(UnaryOperation operation, double[] input)

Methods inherited from class java.lang.Object

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

Methods inherited from interface jdistlib.accelerator.ComputeBackend

automaticRouting, plan, prepareDpotrf, prepareSpotrf, selectedBackend

Methods inherited from interface jdistlib.accelerator.LinearAlgebraBackend

dasum, daxpy, dcopy, dcsrgemm, dcsrmm, dcsrmv, dcsrpotrf, dcsrpotrf, dcsrsv, ddot, dgemm, dgemm, dgemmBatched, dgemv, dgemv, dgeqp3, dger, dgesvd, dgetrf, dgetrfBatched, dnrm2, dpotrf, dpotrfBatched, dscal, dswap, dsyev, dsygvd, dsymm, dsyr, dsyr2, dsyr2k, dsyrk, dsyrk, dsytrf, dtrsm, dtrsm, dtrsv, idamax, prepareDcsr, prepareDcsrpotrf, prepareDcsrpotrf, prepareDge

Methods inherited from interface jdistlib.accelerator.SinglePrecisionLinearAlgebraBackend

isamax, prepareScsr, prepareScsrpotrf, prepareScsrpotrf, prepareSge, sasum, saxpy, scopy, scsrgemm, scsrmm, scsrmv, scsrpotrf, scsrpotrf, scsrsv, sdot, sgemm, sgemm, sgemmBatched, sgemv, sgemv, sgeqp3, sger, sgesvd, sgetrf, sgetrfBatched, snrm2, spotrf, spotrfBatched, sscal, sswap, ssyev, ssygvd, ssymm, ssyr, ssyr2, ssyr2k, ssyrk, ssyrk, ssytrf, strsm, strsm, strsv

Constructor Details

CpuComputeBackend

public CpuComputeBackend()

Method Details

id

public String id()

Specified by:

id in interface ComputeBackend

available

public boolean available()

Specified by:

available in interface ComputeBackend

capabilities

public ComputeCapabilities capabilities()

Specified by:

capabilities in interface ComputeBackend

deviceInfo

public ComputeDeviceInfo deviceInfo()

Description copied from interface: ComputeBackend

Returns stable backend, runtime, driver, and device provenance where available.

Specified by:

deviceInfo in interface ComputeBackend

unary

public double[] unary(UnaryOperation operation,
 double[] input)

Specified by:

unary in interface ComputeBackend

axpy

public double[] axpy(double alpha,
 double[] x,
 double[] y)

Specified by:

axpy in interface ComputeBackend

dot

public double dot(double[] x,
 double[] y)

Specified by:

dot in interface ComputeBackend

matrixMultiply

public double[][] matrixMultiply(double[][] left,
 double[][] right)

Specified by:

matrixMultiply in interface ComputeBackend

prepareTransposeProduct

public PreparedTransposeProduct prepareTransposeProduct(double[][] matrix)

Description copied from interface: ComputeBackend

Keeps a reusable row-by-feature matrix ready for repeated X'v batches.

Specified by:

prepareTransposeProduct in interface ComputeBackend

logisticRegression

public LogisticRegressionBatchResult logisticRegression(double[][] design,
 double[] outcomes,
 double[][] states,
 double priorPrecision)

Specified by:

logisticRegression in interface ComputeBackend

prepareLogisticRegression

public PreparedLogisticRegression prepareLogisticRegression(double[][] design,
 double[] outcomes)

Description copied from interface: ComputeBackend

Keeps reusable data in backend-optimal storage when the backend supports it.

Specified by:

prepareLogisticRegression in interface ComputeBackend

close

public void close()

Specified by:

close in interface AutoCloseable

Specified by:

close in interface ComputeBackend