Physicists at the University of Maryland, US and the quantum computing firm IonQ have found a new way to make a central operation in quantum computing more efficient. By slashing the laser power required to perform a so-called two-qubit gate, the collaborators showed that they could speed up the gate’s operation, thereby boosting the performance of their trapped-ion quantum computer.
The building blocks of a quantum computer are qubits – quantum bits that can be in any superposition of two states. In this work, the researchers used ions as their qubits. Rapidly oscillating electric fields trap the ions in a chain, making it possible to perform computational operations by shining laser light on one or more ions.
Two-qubit entangling gates
These computational operations generally divide into two types: single-qubit gates and two-qubit gates. While single-qubit gates are relatively simple to perform and pose no significant challenges, two-qubit gates cost significant time and power. That has consequences for the overall efficiency of the quantum computer, says Norbert Linke, a fellow of Maryland’s Joint Quantum Institute (JQI) and a co-author of the current study. “The performance of two-qubit entangling gates typically limits the overall system since they require the most calibration time and introduce the most error,” Linke explains. “Improving these gates is therefore crucial to boost the performance and eventually scale up these systems.”
Ideally, gate operations would be fast, use minimal laser power, and leave the qubit in the desired state with no errors (maximum fidelity). In the real world, errors in two-qubit entangling gates come from having imperfect control over experimental parameters such as the frequency of the laser and the trapping field. The general technique to achieve the highest fidelity is therefore to take great care in designing the control signal (that is, the laser beam) that interacts with the ions, eliminating all undesirable effects by fine-tuning the parameters of the protocol. This constrains the design space for the control signal.
Quantum gate teleportation connects atomic qubits in two labs
The IonQ–JQI team’s idea was to sacrifice a small amount of fidelity to save a significant amount of laser power – in some cases an order of magnitude. “We consider the constraints that don’t contribute significantly to the error processes when removed,” explains fellow co-author Yunseong Nam, quantum theory lead at IonQ and adjunct assistant professor at the University of Maryland. “This way, while we sacrifice a minimal amount of fidelity, we can significantly increase the size of the design space, which can then be used to better optimize the power requirement.”
Nam and his colleagues implemented their protocol on the JQI’s programmable trapped-ion quantum hardware with five qubits. When they measured both the power and the fidelity of the gate operations, they found that they could create a maximally entangled state with their method without losing significant fidelity.
Generalizing the technique
Now that the team has carried out a successful proof-of-concept demonstration, its members plan to implement their two-qubit entangling gate in various quantum algorithms. This should allow them to verify whether the newly developed protocol leads to an increase in overall efficiency. Linke adds that they are also exploring ways to generalize their method. “We are working on other schemes for generating entangling gates with different control parameters,” he says. “This will provide the optimal quantum gate mechanism for the particular noise or error characteristics of different devices.”
The paper is published in Physical Review Letters.
