Skip to content

Avoid redundant parity calculation for Pauli amplitude pairs #838

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
whahnsr wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Avoid redundant parity calculation for Pauli amplitude pairs #838

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 5 additions & 4 deletions quest/src/cpu/cpu_subroutines.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -940,7 +940,7 @@ template void cpu_densmatr_allTargDiagMatr_sub<true, false, true, false> (Qure
template <int NumTargs>
INLINE void applyPauliUponAmpPair(
cpu_qcomp* amps, qindex& v, qindex& i0, int* indXY, int& numXY,
qindex& maskXY, qindex& maskYZ, cpu_qcomp& ampFac, cpu_qcomp& pairAmpFac
qindex& maskXY, qindex& maskYZ, int& pairSign, cpu_qcomp& ampFac, cpu_qcomp& pairAmpFac
) {
// This is a subroutine of cpu_statevector_anyCtrlPauliTensorOrGadget_subA() below
// called in a hot-loop (hence it is here inlined) which exists because the caller
Expand All @@ -959,7 +959,7 @@ INLINE void applyPauliUponAmpPair(

// sign of amps due to Y and Z (excludes Y's i factor which is inside pairAmpFac)
int signA = fast_getPlusOrMinusMaskedBitParity(iA, maskYZ);
int signB = fast_getPlusOrMinusMaskedBitParity(iB, maskYZ);
int signB = signA * pairSign;

// mix or swap scaled amp pair (where pairAmpFac includes Y's i factor)
cpu_qcomp ampA = amps[iA];
Expand Down Expand Up @@ -996,6 +996,7 @@ void cpu_statevector_anyCtrlPauliTensorOrGadget_subA(
// prepare masks for extracting Pauli parities
auto maskXY = util_getBitMask(sortedTargsXY);
auto maskYZ = util_getBitMask(util_getConcatenated(y, z));
int pairSign = fast_getPlusOrMinusMaskedBitParity(maskXY, maskYZ);

// use template params to compile-time unroll loops in insertBits() and inner-loop below
int numCtrlBits = accel_tryUseCompileTimeValue<NumCtrls>(ctrls.size());
Expand Down Expand Up @@ -1032,7 +1033,7 @@ void cpu_statevector_anyCtrlPauliTensorOrGadget_subA(
// serial
for (qindex v=0; v<numInnerIts; v++)
applyPauliUponAmpPair<NumTargs>(
amps, v, i0, sortedTargsXY.data(), numTargBits, maskXY, maskYZ, f0, f1);
amps, v, i0, sortedTargsXY.data(), numTargBits, maskXY, maskYZ, pairSign, f0, f1);
}

} else {
Expand All @@ -1047,7 +1048,7 @@ void cpu_statevector_anyCtrlPauliTensorOrGadget_subA(
#pragma omp parallel for
for (qindex v=0; v<numInnerIts; v++)
applyPauliUponAmpPair<NumTargs>(
amps, v, i0, sortedTargsXY.data(), numTargBits, maskXY, maskYZ, f0, f1);
amps, v, i0, sortedTargsXY.data(), numTargBits, maskXY, maskYZ, pairSign, f0, f1);
}
}
}
Expand Down
Loading