Qx.Calc (Qx - Quantum Computing Simulator v0.6.0)

View Source

Shared calculation engine for quantum gate operations.

This module provides the core logic for applying quantum gates to state vectors, used by both Qx.Qubit (single-qubit calculation mode) and Qx.Register (multi-qubit calculation mode).

Design

The module handles:

  • Single-qubit gate application (direct matrix multiplication)
  • Multi-qubit gate application (tensor product expansion)
  • State vector transformations

This centralizes the gate application logic in one place, making it easier to optimize and maintain.

Summary

Functions

apply_cnot(state, control_qubit, target_qubit, num_qubits)

Applies a two-qubit gate (like CNOT) to a state vector.

apply_cswap(state, control, target_a, target_b, num_qubits)

Applies a Fredkin (controlled-SWAP) gate to the quantum state.

apply_single_qubit_gate(state, gate_matrix, target_qubit, num_qubits)

Applies a single-qubit gate to a state vector.

apply_toffoli(state, control1, control2, target, num_qubits)

Applies a Toffoli (CCX) gate to a state vector.

Functions

apply_cnot(state, control_qubit, target_qubit, num_qubits)

@spec apply_cnot(Nx.Tensor.t(), non_neg_integer(), non_neg_integer(), pos_integer()) ::
  Nx.Tensor.t()

Applies a two-qubit gate (like CNOT) to a state vector.

Uses optimized direct statevector manipulation from Qx.CalcFast for improved performance (avoids building 2^n x 2^n matrix).

Parameters

  • state - State vector (Nx tensor)
  • control_qubit - Index of control qubit
  • target_qubit - Index of target qubit
  • num_qubits - Total number of qubits in the system

Examples

iex> reg = Qx.Register.new(2)
iex> Qx.Calc.apply_cnot(reg.state, 0, 1, 2)

apply_cswap(state, control, target_a, target_b, num_qubits)

@spec apply_cswap(
  Nx.Tensor.t(),
  non_neg_integer(),
  non_neg_integer(),
  non_neg_integer(),
  pos_integer()
) :: Nx.Tensor.t()

Applies a Fredkin (controlled-SWAP) gate to the quantum state.

Swaps target_a and target_b when the control qubit is |1⟩.

Parameters

  • state - The quantum state vector
  • control - Index of the control qubit
  • target_a - Index of the first target qubit
  • target_b - Index of the second target qubit
  • num_qubits - Total number of qubits in the system

apply_single_qubit_gate(state, gate_matrix, target_qubit, num_qubits)

@spec apply_single_qubit_gate(
  Nx.Tensor.t(),
  Nx.Tensor.t(),
  non_neg_integer(),
  pos_integer()
) ::
  Nx.Tensor.t()

Applies a single-qubit gate to a state vector.

Uses optimized direct statevector manipulation from Qx.CalcFast for improved performance (10-1000x faster than matrix-based approach).

Parameters

  • state - State vector (Nx tensor)
  • gate_matrix - 2x2 gate matrix (from Qx.Gates)
  • target_qubit - Index of qubit to apply gate to (0-based)
  • num_qubits - Total number of qubits in the system

Examples

# Single qubit (num_qubits = 1, target = 0)
iex> state = Qx.Qubit.new()
iex> gate = Qx.Gates.hadamard()
iex> Qx.Calc.apply_single_qubit_gate(state, gate, 0, 1)

# Multi-qubit register (apply H to qubit 1 in 3-qubit system)
iex> reg = Qx.Register.new(3)
iex> gate = Qx.Gates.hadamard()
iex> Qx.Calc.apply_single_qubit_gate(reg.state, gate, 1, 3)

apply_toffoli(state, control1, control2, target, num_qubits)

@spec apply_toffoli(
  Nx.Tensor.t(),
  non_neg_integer(),
  non_neg_integer(),
  non_neg_integer(),
  pos_integer()
) :: Nx.Tensor.t()

Applies a Toffoli (CCX) gate to a state vector.

Uses optimized direct statevector manipulation from Qx.CalcFast for improved performance (avoids building 2^n x 2^n matrix).

Parameters

  • state - State vector (Nx tensor)
  • control1 - Index of first control qubit
  • control2 - Index of second control qubit
  • target - Index of target qubit
  • num_qubits - Total number of qubits in the system