A Universal Globally Driven Superconducting Quantum Computer

In the rapidly evolving field of quantum computing, the development of scalable architectures is a crucial area of research.

As the quest to increase the number of resilient-to-noise qubits progresses, one of the most pressing issues faced by current quantum computer architectures is the so-called ‘wiring problem’. Each qubit in a quantum computer requires its own control signal in order to operate, leading to a high demand of control lines per qubit which significantly hinders the scalability of quantum processors.

In addition, the efficiency of quantum computer is strongly affected by the presence of unwanted interactions between physical qubits (crosstalks) which must be properly managed in order to reduce the error rate and the ratio of physical-to-logical qubit.

One control line to rule them all

Current quantum processors based on superconducting qubits requires on average 2.5 wires per qubit. Besides, these processors must operate at extremely low temperatures, typically in the millikelvin range, a condition which imposes to cool down each control line (from room) to cryogenic temperatures. This process negatively affects the efficiency of the system, as the control lines transmit thermal noise into the sensitive low-temperature environment hosting the quantum processor. In order to scale from today’s proof of concept to useful quantum computers with thousands of qubits, the ‘wiring’ problem must be properly addressed.

If you are wondering what a quantum computer looks like

Our innovative approach is based on a new and proprietary quantum computing architecture where the number of control lines remains constant (at three), regardless of the number of qubits in the processor. By maintaining a fixed number of control lines, our approach minimizes the introduction of thermal noise, reduces the physical space required to accommodate wiring and enables the design of quantum processors with a vastly greater number of qubits. It unleashes the possibilities to advance quantum computing from experimental setups to practical, large-scale development and applications.

A Turing-like Quantum Computer: the Synchronized Quantum Dance

A Turing machine, devised by Alan Turing in the 1930s, is an abstract mathematical model of computation. It consists of a tape divided into cells, a read/write head that moves left or right along the tape and a finite set of states. The machine follows a set of rules to read from and write onto the tape, enabling it to simulate any algorithmic process through sequential steps of state transitions and tape manipulations.

Similarly, within our architecture logical qubits are designed to move and “dance” across an array of physical qubits. This movement allows the logical qubits to ’emerge’ precisely at strategic locations necessary to execute single and two-qubit gates. The ability to dynamically relocate logical qubits is a fundamental aspect of the model, as it allows qubits to be optimally placed to perform essential quantum operations.

This intricate dance of logical qubits is orchestrated by sending a suitable global pulse sequence. The term ‘global’ is key, as it indicates that every physical qubit in the processor, connected via the same control line, is subject to the same driving pulse. This uniform application of the control pulse across all qubits simplifies the control mechanism, reducing the complexity typically associated with individual qubit manipulation.

It’s an approach that contrasts sharply with traditional methods where each qubit might require separate control lines and dedicated pulse signals. By leveraging a global pulse sequence, we can achieve a more streamlined and scalable control system. This global control not only mitigates the wiring problem but also ensures that all physical qubits respond synchronously to the control signals, enhancing the coherence and reliability of quantum operations.

By integrating this synchronized quantum dance into the architecture, development of systems capable of handling larger and more complex computations with increased efficiency and reliability.

Pictorial representation of a quantum processor model based on our proprietary architecture. The logical qubits are encoded at the column level, represented dark purple stripes. At each time step, the quantum information encoded into logical qubits resides within a specific column of the grid, shown in light purple. By activating global control lines, the interface can be strategically positioned within the processor. To perform a single-qubit gate on a particular qubit, the interface is transferred to the column presenting a crossed purple or green qubit, thus allowing to send a sequence of pulses that rotates the corresponding qubit. An analogous approach is employed for implementing two-qubit gates, by leveraging the crossed red qubit. These qubits act as quantum bridges between rows, enabling entanglement .

Turning a weakness into a strength

In the introduction, we highlighted that current quantum computer architectures are hindered by unwanted interactions between qubits which must be mitigated or accounted for during quantum computations.  Specifically referred to as ‘parasitic ZZ couplings’, these challenging-to-mitigate interactions emerge when two qubits are connected together to perform two-qubit gates and they can lead to unwanted phase accumulations, endangering the reliability of quantum computation.

Unlike traditional approaches, we transform these coupling from an unwanted problem into a functional asset at the of our architecture, forming essential building blocks for our global quantum operation. Specifically, parasitic ZZ couplings are levereged to select a working computational subspace where neighboring qubits are prevented from being in the excited state simultaneously, effectively creating a blockade effect similar to the behavior observed in particles like Rydberg atoms. This enable the above-mentioned ‘quantum dance’ that coordinates qubit interactions.

In addition, this approach eliminates the need to switch the couplings between qubits on and off during computation. Instead, the couplings are set at the fabrication stage and remain constant throughout the computational process

Lastly, the strategic use of parasitic ZZ couplings reduces the need for tunable couplers, which are typically required to manage qubit interactions dynamically. By removing this requirement, our architecture further decreases the number of necessary control lines, contributing to a more streamlined and efficient quantum processor design.

To conclude, our quantum protocol, combined with the fixed couplings, enables us to perform universal quantum computation without the need for independent control of each qubit. The global control pulses we employ ensure that all physical qubits respond uniformly, further simplifying the control mechanism and enhancing the system’s coherence and reliability.