Periodic Reporting for period 2 - SPECTRUM (SuPErConducTing Radio-frequency switch for qUantuM technologies)
Reporting period: 2023-05-01 to 2025-04-30
Despite the high expectations and numerous announcements that have been made over the past ten years, the spread of Quantum Computers (QCs) is still in its infancy. The major factors limiting the diffusion and market penetration of QCs are their low scalability and high cost. Both issues are connected to the bulkiness and complexity of the signal lines that operate the QC. The required large amount of cables undermine the scalability and decrease the thermal stability of the Quantum Processing Units (QPUs). With this project, we aim to increase the scalability and reduce the thermal issues of QPU developing a radiofrequency (RF) switch, QueSt, that allow to simultaneously control the state of multiple qubits through the same cable. QueSt goes well beyond what is achieved with state-of-the-art electronics that typically provide bulky, slowly and energy inefficient solutions. The core component of QueSt is an all-metallic superconducting transistor-controlled via gate voltages. This transistor exploits the peculiar characteristics of a superconducting material to work at frequencies (~1 THz) unachievable with classical semiconductor electronic components and with nearly zero power dissipation.
During SPECTRUM we are going to build a complete test platform QueSt devices. The state-of-the-art nanofabricated prototype of QueSt will be tested in a custom made cryostat able to unleash the true potential of this technology. Furthermore, ultra-fast FPGA-based electronics will take the case of the control of multiple switches, providing an affordable and performant control over the prototype. This platform will be the environment in which QueSt will be studied at strict contact with the state of the art Quantum Processing Units. The experiments performed in real Quantum Computer under the EU-funded Spectrum project will be the first step to the true Quantum Revolution.
During SPECTRUM we are going to build a complete test platform QueSt devices. The state-of-the-art nanofabricated prototype of QueSt will be tested in a custom made cryostat able to unleash the true potential of this technology. Furthermore, ultra-fast FPGA-based electronics will take the case of the control of multiple switches, providing an affordable and performant control over the prototype. This platform will be the environment in which QueSt will be studied at strict contact with the state of the art Quantum Processing Units. The experiments performed in real Quantum Computer under the EU-funded Spectrum project will be the first step to the true Quantum Revolution.
The Spectrum Consortium has worked out a novel solution that is aimed at enhancing the control over superconducting Qubits while reducing the amount of lines inside the cryostat, which is in fact QueSt, i.e. a Quantum Superconducting Switch. QueSt is a nanofabricated network switch based on a superconducting high-frequency transistor, controlled by gate voltages.
Starting from the original product concept of fully metallic device, the team has been trying different solutions by varying materials and circuitry designs, achieving an inexpensive and easily scalable product.
Through an iterative trials and errors approach, different materials combinations realized through sputtering and other nanodeposition techniques.
In this way, a novel material platform was developed, based on InAsOI architecture. The approach leverages accurate epitaxial growth and selective doping techniques to precisely tune the electrical characteristics of the semiconductor layer. When combined with a conventional superconductive film (defining the drain and source) and a MOS-like gate structure, the team was able to address critical performance challenges in high-speed switching and quantum processor applications by leveraging Josephson Junction effect.
As a result of the above activity, the newly discovered and patented InAsOI platform is highlighted as a more robust alternative to the metallic approach, with structural similarities to silicon-on-insulator (SOI) systems. In particular, InAsOI features an InAs semiconductor layer atop a cryogenic insulator that is capable of providing several advantages.
Firstly, it enables stable superconductivity due to the proximity effect which, when coupled with superconducting leads, ensures a controlled, non-dissipative current flow, thereby reducing overheating risks and maintaining signal integrity. This translates into higher stability and performance, as InAsOI can operate at lower gate voltages and supports higher critical current densities, both essential for reliable superconducting operations. Furthermore, it provides predictable gate control, allowing for stable, low-voltage control over the conduction properties, which minimizes overheating issues and enhances compatibility with cryogenic conditions.
Within this framework, the CNR team has successfully fabricated various samples of JoFETs (Josephson Junction Field-Effect Transistor) utilizing this material platform. The findings indicate that in these JoFETs, gate-controlled modulation precisely manages superconducting pathways. This results in more stable gate control, as InAsOI FETs operate effectively under lower gate voltages, leading to improved reliability and energy efficiency. Moreover, the platform demonstrates scalability for quantum applications, given that its low energy dissipation and high critical current density make it highly suitable for integration into quantum computing systems, where superconducting efficiency is paramount.
Despite great promises, the workplan was not fulfilled because of the delays with delivery of some essential components.
As a consequence, the pilot study at CUT was performed in a partial manner.
Nevertheless, the results shown during the even partial pilot study has confirmed the ability of QueSt as a superconductive cryocomponent not generating any heat effect during functioning. This is a ground breaking achievement as all the state of the art switches produce some heat within the cryostat, thus creating noise and impeding smooth measurements of the Qubits.
Starting from the original product concept of fully metallic device, the team has been trying different solutions by varying materials and circuitry designs, achieving an inexpensive and easily scalable product.
Through an iterative trials and errors approach, different materials combinations realized through sputtering and other nanodeposition techniques.
In this way, a novel material platform was developed, based on InAsOI architecture. The approach leverages accurate epitaxial growth and selective doping techniques to precisely tune the electrical characteristics of the semiconductor layer. When combined with a conventional superconductive film (defining the drain and source) and a MOS-like gate structure, the team was able to address critical performance challenges in high-speed switching and quantum processor applications by leveraging Josephson Junction effect.
As a result of the above activity, the newly discovered and patented InAsOI platform is highlighted as a more robust alternative to the metallic approach, with structural similarities to silicon-on-insulator (SOI) systems. In particular, InAsOI features an InAs semiconductor layer atop a cryogenic insulator that is capable of providing several advantages.
Firstly, it enables stable superconductivity due to the proximity effect which, when coupled with superconducting leads, ensures a controlled, non-dissipative current flow, thereby reducing overheating risks and maintaining signal integrity. This translates into higher stability and performance, as InAsOI can operate at lower gate voltages and supports higher critical current densities, both essential for reliable superconducting operations. Furthermore, it provides predictable gate control, allowing for stable, low-voltage control over the conduction properties, which minimizes overheating issues and enhances compatibility with cryogenic conditions.
Within this framework, the CNR team has successfully fabricated various samples of JoFETs (Josephson Junction Field-Effect Transistor) utilizing this material platform. The findings indicate that in these JoFETs, gate-controlled modulation precisely manages superconducting pathways. This results in more stable gate control, as InAsOI FETs operate effectively under lower gate voltages, leading to improved reliability and energy efficiency. Moreover, the platform demonstrates scalability for quantum applications, given that its low energy dissipation and high critical current density make it highly suitable for integration into quantum computing systems, where superconducting efficiency is paramount.
Despite great promises, the workplan was not fulfilled because of the delays with delivery of some essential components.
As a consequence, the pilot study at CUT was performed in a partial manner.
Nevertheless, the results shown during the even partial pilot study has confirmed the ability of QueSt as a superconductive cryocomponent not generating any heat effect during functioning. This is a ground breaking achievement as all the state of the art switches produce some heat within the cryostat, thus creating noise and impeding smooth measurements of the Qubits.
Recognizing the novel structure and superconductive properties of InAsOI, the team has identified key added value propositions for customers in the quantum sector. These include Innovations in Material Composition and Layering, where the specific layering, particularly the InAlAs intermediate layer, enables reliable control of superconductive properties, setting InAsOI apart from previous superconductive designs. Additionally, it offers enhanced device performance, as adjustments to the critical current density and controlled decoupling between junctions support advanced quantum applications, making InAsOI an attractive option for commercialization in superconducting electronics.
All the above features, jointly make the InAsOl platform much more suitable for the implementation within quantum tech than the originally ideated full metallic solution. Last but not least, the materials used in this solution are well established, compatible with any chip production technology and thus enable a flexible integration into the QPU chip, independently from the material platform used by the QPU manufacturer.
To summarize, we realized a very sensitive measurement that shows no heating at all from a QueSt prototype switching event. This supports its suitability for thermally sensitive applications such as qubit readout isolation.
Within the project, dynamically configurable control of a QPU and on the multiplexed characterization of several QPUs was conducted. The results of experiments in a quantum set up at CUT demonstrate that serialization overhead remains manageable across all tested configurations. For the most aggressive multiplexing scenario (all 13 qubits sharing a single control line), the average serialization overhead was 0.13 μs, representing a small fraction of the mean routing overhead of 2.43 μs. This significant difference highlights the viability of time-multiplexing as a control line reduction strategy.
All the above features, jointly make the InAsOl platform much more suitable for the implementation within quantum tech than the originally ideated full metallic solution. Last but not least, the materials used in this solution are well established, compatible with any chip production technology and thus enable a flexible integration into the QPU chip, independently from the material platform used by the QPU manufacturer.
To summarize, we realized a very sensitive measurement that shows no heating at all from a QueSt prototype switching event. This supports its suitability for thermally sensitive applications such as qubit readout isolation.
Within the project, dynamically configurable control of a QPU and on the multiplexed characterization of several QPUs was conducted. The results of experiments in a quantum set up at CUT demonstrate that serialization overhead remains manageable across all tested configurations. For the most aggressive multiplexing scenario (all 13 qubits sharing a single control line), the average serialization overhead was 0.13 μs, representing a small fraction of the mean routing overhead of 2.43 μs. This significant difference highlights the viability of time-multiplexing as a control line reduction strategy.