Header file for AVX-optimized implementations for applying multi-qubit gate kernels to quantum state vectors and matrices. More...

#include <immintrin.h>
#include "matrix.h"
#include "matrix_float.h"
#include "common.h"
#include "apply_large_kernel_to_input.h"

Include dependency graph for apply_large_kernel_to_input_AVX.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Functions
void	apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP (Matrix &two_qbit_unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)
	Apply two-qubit gate kernel to an input matrix using AVX optimization and OpenMP parallelization. More...

void	apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP32 (Matrix_float &two_qbit_unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_AVX (Matrix &u3_1qbit1, Matrix &u3_qbit2, Matrix &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)
	Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization. More...

void	apply_crot_kernel_to_matrix_input_AVX32 (Matrix_float &u3_1qbit1, Matrix_float &u3_qbit2, Matrix_float &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_AVX_parallel (Matrix &u3_1qbit1, Matrix &u3_1qbit2, Matrix &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)
	Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization and parallel processing. More...

void	apply_crot_kernel_to_matrix_input_AVX_parallel32 (Matrix_float &u3_1qbit1, Matrix_float &u3_1qbit2, Matrix_float &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_from_right_AVX (Matrix &u3_1qbit1, Matrix &u3_qbit2, Matrix &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_from_right_AVX32 (Matrix_float &u3_1qbit1, Matrix_float &u3_qbit2, Matrix_float &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_from_right_AVX_parallel (Matrix &u3_1qbit1, Matrix &u3_1qbit2, Matrix &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_crot_kernel_to_matrix_input_from_right_AVX_parallel32 (Matrix_float &u3_1qbit1, Matrix_float &u3_1qbit2, Matrix_float &input, const int &target_qbit, const int &control_qbit, const int &matrix_size)

void	apply_large_kernel_from_right_AVX (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_from_right_AVX32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_from_right_AVX_OpenMP (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_from_right_AVX_OpenMP32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_from_right_AVX_TBB (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_from_right_AVX_TBB32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_to_input_AVX (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)
	Apply multi-qubit gate kernel to an input matrix using AVX optimization. More...

void	apply_large_kernel_to_input_AVX32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_to_input_AVX_OpenMP (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)
	Apply multi-qubit gate kernel to an input matrix using AVX optimization and OpenMP parallelization. More...

void	apply_large_kernel_to_input_AVX_OpenMP32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

void	apply_large_kernel_to_input_AVX_TBB (Matrix &unitary, Matrix &input, std::vector< int > involved_qbits, const int &matrix_size)
	Apply multi-qubit gate kernel to an input matrix using AVX optimization and TBB parallelization. More...

void	apply_large_kernel_to_input_AVX_TBB32 (Matrix_float &unitary, Matrix_float &input, std::vector< int > involved_qbits, const int &matrix_size)

__m256d	complex_mult_AVX (__m256d input_vec, __m256d unitary_row_vec, __m256d neg)
	Perform complex multiplication using AVX intrinsics. More...

__m256d *	construct_mv_xy_vectors (const Matrix &gate_kernel_unitary, const int &matrix_size)
	Precompute AVX vectors for the unitary matrix to optimize complex multiplication. More...

__m256d	get_AVX_vector (double element_outer, double element_inner)
	Helper function to load and prepare AVX vectors with outer and inner elements for complex multiplication. More...

void	get_block_indices (int N, const std::vector< int > &tq, const std::vector< int > &non_targets, int iter_idx, std::vector< int > &indices)
	Compute the indices for a given block in the quantum state. More...

void	get_block_indices_fast (int iter_idx, const std::vector< int > &target_qubits, const std::vector< int > &non_targets, const std::vector< int > &block_pattern, std::vector< int > &indices)
	Efficiently compute the indices for a given block using precomputed patterns. More...

void	precompute_index_mapping (const std::vector< int > &target_qubits, const std::vector< int > &non_targets, std::vector< int > &block_pattern)
	Precompute the index mapping for target and non-target qubits. More...

Detailed Description

Header file for AVX-optimized implementations for applying multi-qubit gate kernels to quantum state vectors and matrices.

Definition in file apply_large_kernel_to_input_AVX.h.

Function Documentation

◆ apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP()

void apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP	(	Matrix &	two_qbit_unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Apply two-qubit gate kernel to an input matrix using AVX optimization and OpenMP parallelization.

Parameters

two_qbit_unitary	The 4x4 unitary matrix representing the two-qubit gate operation
input	The input matrix on which the transformation is applied
involved_qbits	The qubits involved in the gate operation (2 qubits)
matrix_size	The size of the input matrix (should be a power of 2)

Definition at line 3520 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP32()

void apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP32	(	Matrix_float &	two_qbit_unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5960 of file apply_large_kernel_to_input_AVX.cpp.

◆ apply_crot_kernel_to_matrix_input_AVX()

void apply_crot_kernel_to_matrix_input_AVX	(	Matrix &	u3_1qbit1,
		Matrix &	u3_1qbit2,
		Matrix &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization.

Parameters

u3_1qbit1	The U3 gate matrix for the first qubit
u3_qbit2	The U3 gate matrix for the second qubit
input	The input matrix on which the transformation is applied
target_qbit	The target qubit index
control_qbit	The control qubit index
matrix_size	The size of the input matrix (should be a power of 2)

Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization.

Parameters

u3_1qbit1	The 2x2 kernel to be applied on target \|1>
u3_1qbit2	The 2x2 kernel to be applied on target \|0>
input	The input matrix on which the transformation is applied
target_qbit	The target qubit
control_qbit	The control qubit
matrix_size	The size of the input

Definition at line 3617 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_AVX32()

void apply_crot_kernel_to_matrix_input_AVX32	(	Matrix_float &	u3_1qbit1,
		Matrix_float &	u3_qbit2,
		Matrix_float &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 6160 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_AVX_parallel()

void apply_crot_kernel_to_matrix_input_AVX_parallel	(	Matrix &	u3_1qbit1,
		Matrix &	u3_1qbit2,
		Matrix &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization and parallel processing.

Parameters

u3_1qbit1	The U3 gate matrix for the first qubit
u3_1qbit2	The U3 gate matrix for the second qubit
input	The input matrix on which the transformation is applied
target_qbit	The target qubit index
control_qbit	The control qubit index
matrix_size	The size of the input matrix (should be a power of 2)

Apply controlled rotation (CROT) kernel to a matrix input using AVX optimization and parallel processing.

Parameters

u3_1qbit1	The 2x2 kernel to be applied on target \|1>
u3_1qbit2	The 2x2 kernel to be applied on target \|0>
input	The input matrix on which the transformation is applied
target_qbit	The target qubit
control_qbit	The control qubit
matrix_size	The size of the input

Definition at line 3836 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_AVX_parallel32()

void apply_crot_kernel_to_matrix_input_AVX_parallel32	(	Matrix_float &	u3_1qbit1,
		Matrix_float &	u3_1qbit2,
		Matrix_float &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 6000 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_from_right_AVX()

void apply_crot_kernel_to_matrix_input_from_right_AVX	(	Matrix &	u3_1qbit1,
		Matrix &	u3_qbit2,
		Matrix &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 4089 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_from_right_AVX32()

void apply_crot_kernel_to_matrix_input_from_right_AVX32	(	Matrix_float &	u3_1qbit1,
		Matrix_float &	u3_qbit2,
		Matrix_float &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 6203 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_from_right_AVX_parallel()

void apply_crot_kernel_to_matrix_input_from_right_AVX_parallel	(	Matrix &	u3_1qbit1,
		Matrix &	u3_1qbit2,
		Matrix &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 4217 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_crot_kernel_to_matrix_input_from_right_AVX_parallel32()

void apply_crot_kernel_to_matrix_input_from_right_AVX_parallel32	(	Matrix_float &	u3_1qbit1,
		Matrix_float &	u3_1qbit2,
		Matrix_float &	input,
		const int &	target_qbit,
		const int &	control_qbit,
		const int &	matrix_size
	)

Definition at line 6060 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ apply_large_kernel_from_right_AVX()

void apply_large_kernel_from_right_AVX	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5568 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_from_right_AVX32()

void apply_large_kernel_from_right_AVX32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5610 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_from_right_AVX_OpenMP()

void apply_large_kernel_from_right_AVX_OpenMP	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5582 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

◆ apply_large_kernel_from_right_AVX_OpenMP32()

void apply_large_kernel_from_right_AVX_OpenMP32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5624 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

◆ apply_large_kernel_from_right_AVX_TBB()

void apply_large_kernel_from_right_AVX_TBB	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5596 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_from_right_AVX_TBB32()

void apply_large_kernel_from_right_AVX_TBB32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 5638 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX()

void apply_large_kernel_to_input_AVX	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Apply multi-qubit gate kernel to an input matrix using AVX optimization.

Parameters

unitary	The 2^Nx2^N unitary matrix representing the gate operation
input	The input matrix on which the transformation is applied
involved_qbits	The qubits involved in the gate operation
matrix_size	The size of the input matrix (should be a power of 2)

Definition at line 90 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX32()

void apply_large_kernel_to_input_AVX32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 4345 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX_OpenMP()

void apply_large_kernel_to_input_AVX_OpenMP	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Apply multi-qubit gate kernel to an input matrix using AVX optimization and OpenMP parallelization.

Parameters

unitary	The 2^Nx2^N unitary matrix representing the gate operation
input	The input matrix on which the transformation is applied
involved_qbits	The qubits involved in the gate operation
matrix_size	The size of the input matrix (should be a power of 2)

Definition at line 145 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX_OpenMP32()

void apply_large_kernel_to_input_AVX_OpenMP32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 4389 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX_TBB()

void apply_large_kernel_to_input_AVX_TBB	(	Matrix &	unitary,
		Matrix &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Apply multi-qubit gate kernel to an input matrix using AVX optimization and TBB parallelization.

Parameters

unitary	The 2^Nx2^N unitary matrix representing the gate operation
input	The input matrix on which the transformation is applied
involved_qbits	The qubits involved in the gate operation
matrix_size	The size of the input matrix (should be a power of 2)

Definition at line 373 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ apply_large_kernel_to_input_AVX_TBB32()

void apply_large_kernel_to_input_AVX_TBB32	(	Matrix_float &	unitary,
		Matrix_float &	input,
		std::vector< int >	involved_qbits,
		const int &	matrix_size
	)

Definition at line 4433 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ complex_mult_AVX()

__m256d complex_mult_AVX	(	__m256d	input_vec,
		__m256d	unitary_row_vec,
		__m256d	neg
	)

inline

Perform complex multiplication using AVX intrinsics.

Parameters

input_vec	AVX vector containing the input complex number (real and imaginary parts)
unitary_row_vec	AVX vector containing the unitary matrix row element (real and imaginary parts)
neg	AVX vector containing sign pattern for complex multiplication

Returns: AVX vector containing the result of complex multiplication

Definition at line 62 of file apply_large_kernel_to_input_AVX.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ construct_mv_xy_vectors()

__m256d* construct_mv_xy_vectors	(	const Matrix &	gate_kernel_unitary,
		const int &	matrix_size
	)

inline

Precompute AVX vectors for the unitary matrix to optimize complex multiplication.

Parameters

gate_kernel_unitary	The unitary matrix of the gate operation
matrix_size	The size of the unitary matrix

Returns: Pointer to the precomputed AVX vectors

Precompute AVX vectors for the unitary matrix to optimize complex multiplication.

Parameters

input	The input matrix to be updated
new_block_real	The real parts of the new block
new_block_imag	The imaginary parts of the new block
indices	The indices where the new block should be written Precompute AVX vectors for the unitary matrix to optimize complex multiplication
gate_kernel_unitary	The unitary matrix of the gate operation
matrix_size	The size of the unitary matrix

Returns: Pointer to the precomputed AVX vectors

Definition at line 481 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ get_AVX_vector()

__m256d get_AVX_vector	(	double *	element_outer,
		double *	element_inner
	)

inline

Helper function to load and prepare AVX vectors with outer and inner elements for complex multiplication.

Parameters

element_outer	Pointer to the outer element (real and imaginary parts)
element_inner	Pointer to the inner element (real and imaginary parts)

Returns: Prepared AVX vector for complex multiplication

Definition at line 43 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ get_block_indices()

void get_block_indices	(	int	N,
		const std::vector< int > &	tq,
		const std::vector< int > &	non_targets,
		int	iter_idx,
		std::vector< int > &	indices
	)

inline

Compute the indices for a given block in the quantum state.

Parameters

N	The number of qubits
tq	The target qubits involved in the gate operation
non_targets	The qubits not involved in the gate operation
iter_idx	The current iteration index
indices	The computed indices for the current block (output parameter)

◆ get_block_indices_fast()

void get_block_indices_fast	(	int	iter_idx,
		const std::vector< int > &	target_qubits,
		const std::vector< int > &	non_targets,
		const std::vector< int > &	block_pattern,
		std::vector< int > &	indices
	)

inline

Efficiently compute the indices for a given block using precomputed patterns.

Parameters

iter_idx	The current iteration index
target_qubits	The qubits involved in the gate operation
non_targets	The qubits not involved in the gate operation
block_pattern	The precomputed index mapping
indices	The computed indices for the current block (output parameter)
iter_idx	The current iteration index
target_qubits	The qubits involved in the gate operation
non_targets	The qubits not involved in the gate operation
block_pattern	The precomputed index mapping
indices	The computed indices for the current block

Definition at line 451 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

◆ precompute_index_mapping()

void precompute_index_mapping	(	const std::vector< int > &	target_qubits,
		const std::vector< int > &	non_targets,
		std::vector< int > &	block_pattern
	)

Precompute the index mapping for target and non-target qubits.

Parameters

target_qubits	The qubits involved in the gate operation
non_targets	The qubits not involved in the gate operation
block_pattern	The precomputed index mapping (output parameter)

Definition at line 426 of file apply_large_kernel_to_input_AVX.cpp.

Here is the caller graph for this function:

Functions

Detailed Description

Function Documentation

◆ apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP()

◆ apply_2qbit_kernel_to_matrix_input_parallel_AVX_OpenMP32()

◆ apply_crot_kernel_to_matrix_input_AVX()

◆ apply_crot_kernel_to_matrix_input_AVX32()

◆ apply_crot_kernel_to_matrix_input_AVX_parallel()

◆ apply_crot_kernel_to_matrix_input_AVX_parallel32()

◆ apply_crot_kernel_to_matrix_input_from_right_AVX()

◆ apply_crot_kernel_to_matrix_input_from_right_AVX32()

◆ apply_crot_kernel_to_matrix_input_from_right_AVX_parallel()

◆ apply_crot_kernel_to_matrix_input_from_right_AVX_parallel32()

◆ apply_large_kernel_from_right_AVX()

◆ apply_large_kernel_from_right_AVX32()

◆ apply_large_kernel_from_right_AVX_OpenMP()

◆ apply_large_kernel_from_right_AVX_OpenMP32()

◆ apply_large_kernel_from_right_AVX_TBB()

◆ apply_large_kernel_from_right_AVX_TBB32()

◆ apply_large_kernel_to_input_AVX()

◆ apply_large_kernel_to_input_AVX32()

◆ apply_large_kernel_to_input_AVX_OpenMP()

◆ apply_large_kernel_to_input_AVX_OpenMP32()

◆ apply_large_kernel_to_input_AVX_TBB()

◆ apply_large_kernel_to_input_AVX_TBB32()

◆ complex_mult_AVX()

◆ construct_mv_xy_vectors()

◆ get_AVX_vector()

◆ get_block_indices()

◆ get_block_indices_fast()

◆ precompute_index_mapping()