Package ffx.numerics.multipole

Class CoulombTensorQISIMD

java.lang.Object

ffx.numerics.multipole.MultipoleTensorSIMD

ffx.numerics.multipole.CoulombTensorQISIMD

Direct Known Subclasses:

EwaldTensorQISIMD, GKTensorQISIMD, TholeTensorQISIMD

public class CoulombTensorQISIMD
extends MultipoleTensorSIMD

The CoulombTensorQISIMD class computes derivatives of 1/|r| via recursion to arbitrary order for Cartesian multipoles in a quasi-internal frame using SIMD instructions.

Since:

1.0

Author:

Michael J. Schnieders

See Also:

• Matt Challacombe, Eric Schwegler and Jan Almlof, Modern developments in Hartree-Fock theory: Fast methods for computing the Coulomb matrix. Computational Chemistry: Review of Current Trends. pp. 53-107, Ed. J. Leczszynski, World Scientifc, 1996.

Field Summary

Fields inherited from class ffx.numerics.multipole.MultipoleTensorSIMD

coordinates, coulombSource, E000, E001, E002, E003, E010, E011, E012, E020, E021, E030, E100, E101, E102, E110, E111, E120, E200, E201, E210, E300, il, im, in, o1, operator, order, R, R000, R001, R002, R003, R004, R005, R006, R010, R011, R012, R013, R014, R015, R020, R021, R022, R023, R024, R030, R031, R032, R033, R040, R041, R042, R050, R051, R060, R100, R101, R102, R103, R104, R105, R110, R111, R112, R113, R114, R120, R121, R122, R123, R130, R131, R132, R140, R141, R150, r2, R200, R201, R202, R203, R204, R210, R211, R212, R213, R220, R221, R222, R230, R231, R240, R300, R301, R302, R303, R310, R311, R312, R320, R321, R330, R400, R401, R402, R410, R411, R420, R500, R501, R510, R600, size, t000, t001, t002, t003, t004, t005, t006, t010, t011, t012, t013, t014, t015, t020, t021, t022, t023, t024, t030, t031, t032, t033, t040, t041, t042, t050, t051, t060, t100, t101, t102, t103, t104, t105, t110, t111, t112, t113, t114, t120, t121, t122, t123, t130, t131, t132, t140, t141, t150, t200, t201, t202, t203, t204, t210, t211, t212, t213, t220, t221, t222, t230, t231, t240, t300, t301, t302, t303, t310, t311, t312, t320, t321, t330, t400, t401, t402, t410, t411, t420, t500, t501, t510, t600, work, x, y, z

Constructor Summary

Constructors

Constructor	Description
CoulombTensorQISIMD(int order)	Create a new CoulombTensorQI object.

Method Summary

Modifier and Type	Method	Description
protected void	chargeIPotentialAtK(PolarizableMultipoleSIMD mI, int order)	Compute the field components due to site I charge at site K.
protected void	chargeKPotentialAtI(PolarizableMultipoleSIMD mK, int order)	Compute the field components due to site K charge at site I.
protected void	dipoleIPotentialAtK(DoubleVector uxi, DoubleVector uyi, DoubleVector uzi, int order)	Compute the induced dipole field components due to site I at site K.
protected void	dipoleKPotentialAtI(DoubleVector uxk, DoubleVector uyk, DoubleVector uzk, int order)	Compute the induced dipole field components due to site K at site I.
protected void	multipoleIPotentialAtK(PolarizableMultipoleSIMD mI, int order)
protected void	multipoleKPotentialAtI(PolarizableMultipoleSIMD mK, int order)
protected void	order1()	Hard coded tensor computation up to 1st order.
protected void	order2()	Hard coded tensor computation up to 2nd order.
protected void	order3()	Hard coded tensor computation up to 3rd order.
protected void	order4()	Hard coded tensor computation up to 4th order.
protected void	order5()	Hard coded tensor computation up to 5th order.
protected void	order6()	Hard coded tensor computation up to 6th order.
protected void	quadrupoleIPotentialAtK(PolarizableMultipoleSIMD mI, int order)	Compute the field components due to site I quadrupole at site K.
protected void	quadrupoleKPotentialAtI(PolarizableMultipoleSIMD mK, int order)	Compute the field components due to site K quadrupole at site I.
void	setR(DoubleVector dx, DoubleVector dy, DoubleVector dz)	Set the separation vector.
protected void	source(DoubleVector[] T000)	Generate source terms for the Coulomb Challacombe et al. recursion.

Methods inherited from class ffx.numerics.multipole.MultipoleTensorSIMD

dipoleTorque, generateTensor, getSource, multipoleEnergy, multipoleEnergy, multipoleEnergyAndGradient, multipoleGradient, multipoleTorque, polarizationEnergy, polarizationEnergyAndGradient, polarizationEnergyS, quadrupoleTorque, setR, ti

Methods inherited from class java.lang.Object

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

Constructor Details

CoulombTensorQISIMD

public CoulombTensorQISIMD(int order)

Create a new CoulombTensorQI object.

Parameters:

order - The tensor order.

Method Details

setR

public void setR(DoubleVector dx,
 DoubleVector dy,
 DoubleVector dz)

Set the separation vector.

Specified by:

setR in class MultipoleTensorSIMD

Parameters:

dx - Separation along the X-axis.

dy - Separation along the Y-axis.

dz - Separation along the Z-axis.

source

protected void source(DoubleVector[] T000)

Generate source terms for the Coulomb Challacombe et al. recursion.

Specified by:

source in class MultipoleTensorSIMD

Parameters:

T000 - Location to store the source terms.

order1

protected void order1()

Hard coded tensor computation up to 1st order. This code is auto-generated for both Global and QI frames.

Specified by:

order1 in class MultipoleTensorSIMD

order2

protected void order2()

Hard coded tensor computation up to 2nd order. This code is auto-generated for both Global and QI frames.

Specified by:

order2 in class MultipoleTensorSIMD

order3

protected void order3()

Hard coded tensor computation up to 3rd order. This code is auto-generated for both Global and QI frames.

Specified by:

order3 in class MultipoleTensorSIMD

order4

protected void order4()

Hard coded tensor computation up to 4th order. This code is auto-generated for both Global and QI frames.

Specified by:

order4 in class MultipoleTensorSIMD

order5

protected void order5()

Hard coded tensor computation up to 5th order. This code is auto-generated for both Global and QI frames.
The 5th order recursion is needed for quadrupole-quadrupole forces.

Specified by:

order5 in class MultipoleTensorSIMD

order6

protected void order6()

Hard coded tensor computation up to 6th order. This code is auto-generated for both Global and QI frames.
This is needed for quadrupole-quadrupole forces and orthogonal space sampling.

Specified by:

order6 in class MultipoleTensorSIMD

multipoleKPotentialAtI

protected void multipoleKPotentialAtI(PolarizableMultipoleSIMD mK,
 int order)

Specified by:

multipoleKPotentialAtI in class MultipoleTensorSIMD

chargeKPotentialAtI

protected void chargeKPotentialAtI(PolarizableMultipoleSIMD mK,
 int order)

Compute the field components due to site K charge at site I.

Specified by:

chargeKPotentialAtI in class MultipoleTensorSIMD

Parameters:

mK - MultipoleTensorSIMD at site K.

order - Potential order.

dipoleKPotentialAtI

protected void dipoleKPotentialAtI(DoubleVector uxk,
 DoubleVector uyk,
 DoubleVector uzk,
 int order)

Compute the induced dipole field components due to site K at site I.

Specified by:

dipoleKPotentialAtI in class MultipoleTensorSIMD

Parameters:

uxk - X-dipole component.

uyk - Y-dipole component.

uzk - Z-dipole component.

order - Potential order.

quadrupoleKPotentialAtI

protected void quadrupoleKPotentialAtI(PolarizableMultipoleSIMD mK,
 int order)

Compute the field components due to site K quadrupole at site I.

Specified by:

quadrupoleKPotentialAtI in class MultipoleTensorSIMD

Parameters:

mK - MultipoleTensorSIMD at site K.

order - Potential order.

multipoleIPotentialAtK

protected void multipoleIPotentialAtK(PolarizableMultipoleSIMD mI,
 int order)

Specified by:

multipoleIPotentialAtK in class MultipoleTensorSIMD

chargeIPotentialAtK

protected void chargeIPotentialAtK(PolarizableMultipoleSIMD mI,
 int order)

Compute the field components due to site I charge at site K.

Specified by:

chargeIPotentialAtK in class MultipoleTensorSIMD

Parameters:

mI - MultipoleTensorSIMD at site I.

order - Potential order.

dipoleIPotentialAtK

protected void dipoleIPotentialAtK(DoubleVector uxi,
 DoubleVector uyi,
 DoubleVector uzi,
 int order)

Compute the induced dipole field components due to site I at site K.

Specified by:

dipoleIPotentialAtK in class MultipoleTensorSIMD

Parameters:

uxi - X-dipole component.

uyi - Y-dipole component.

uzi - Z-dipole component.

order - Potential order.

quadrupoleIPotentialAtK

protected void quadrupoleIPotentialAtK(PolarizableMultipoleSIMD mI,
 int order)

Compute the field components due to site I quadrupole at site K.

Specified by:

quadrupoleIPotentialAtK in class MultipoleTensorSIMD

Parameters:

mI - MultipoleTensorSIMD at site I.

order - Potential order.