Qubits vs Qudits: Energy Efficiency in Quantum State Preparation Explored

Quantum information processing can be performed using either qubits or qudits, with the choice impacting the efficiency of the process. A study by researchers at the Universidade Estadual de Campinas has found that using qubits requires more energy than using qudits for quantum state preparation in photonic circuits. However, circuits with dedicated Controlled NOT (CNOT) blocks have smaller energy consumption, making them viable even at large qubit numbers. The research highlights the importance of considering energy consumption when scaling up quantum integrated photonic devices and the need for a more detailed evaluation of the resource costs associated with different quantum information processing approaches.

What is the Energy Cost of Quantum State Preparation in Photonic Circuits?

Quantum information processing, which involves the manipulation of quantum states, can be performed using either qubits or qudits. Qubits and qudits are units of quantum information, with qudits being a generalization of qubits to higher dimensions. The choice between qubits and qudits can impact the efficiency of quantum information processing in terms of the number of photons and waveguides required, as well as the electrical energy cost.

Qudits are more efficient in terms of requiring fewer photons, in principle only one, to encode the state. On the other hand, qubits are more efficient in terms of the number of waveguides used to reach the same dimension as qudits. Qubits require 2 log2d waveguides, while qudits need d waveguides. For tasks requiring large dimensions, this would suggest that qubits are the better option, as other resources scale at least polynomially with the number of waveguides in the photonic integrated circuit (PIC).

How Does the Task of Quantum State Preparation Impact the Choice Between Qubits and Qudits?

The task of quantum state preparation, which is relevant for variational quantum algorithms, requires the circuit to have a number of Controlled NOT (CNOT) gates that is exponential in the number of qubits. Both qubit and qudit implementations suffer from an exponential resource cost, necessitating a more detailed evaluation.

When comparing the qubit and qudit approaches in terms of the amount of electrical energy that must be spent on average to program a photonic circuit to perform quantum state preparation, it was found that using qubits requires more energy than using qudits if a PIC with a fully reconfigurable array of interferometers is used. However, a circuit with dedicated CNOT blocks has a much smaller energy consumption, remaining viable even at large qubit numbers where more important bottlenecks, such as the probabilistic nature of the CNOT gates, come into play.

What is the Role of Photonic Integrated Circuits in Quantum Information Processing?

Since 2001, quantum information processing using photons in the discrete variable paradigm has moved from the optical table to photonic integrated circuits (PICs). This transition accelerated as high-quality lithography equipment and photonic foundry services became available, allowing an increase in device size, complexity, and fabrication throughput.

In the past decade, progress has been made in both gate and measurement-based quantum devices for quantum information processing, quantum state preparation, and quantum simulation tasks, albeit still at a small scale in comparison to the hundreds of qubits now available in other platforms. Similar progress has been made with devices which use continuous-variable and hybrid encodings of quantum information.

How Does Energy Consumption Impact the Scalability of Quantum Integrated Photonic Devices?

As the size of the circuits scales up, the energy consumption of the electro-optical modulators (EOMs) used in each Mach-Zehnder Interferometer (MZI) becomes a relevant issue. This is the main question addressed in the study by Maria Carolina Volpato and Pierre-Louis de Assis from the Gleb Wataghin Institute of Physics at the Universidade Estadual de Campinas.

To perform a quantitative evaluation of the energy consumption of quantum PICs, the researchers constrained their analysis to the task of arbitrary quantum state preparation. They investigated the case of path-encoded qubits and qudits. Even though qudits require exponentially more waveguides to implement a state of a given dimension than qubits, they were used as a benchmark.

What are the Implications of the Study for Quantum Information Processing?

The study shows that quantum state preparation using qubits can be even more energy-intensive than when using qudits if a programmable circuit is used, even when considering state-of-the-art EOMs. However, feasible gate-based quantum state preparation in integrated photonic circuits can be achieved using qubits, but must rely on fixed circuits that keep a minimal degree of tunability to compensate for fabrication errors.

This research provides valuable insights into the energy cost of quantum state preparation using qubits and qudits in integrated photonic circuits. It highlights the need for a more detailed evaluation of the resource costs associated with different quantum information processing approaches, and the importance of considering energy consumption when scaling up quantum integrated photonic devices.