‘Shape-shifting’ quantum computer uses light to switch between different tasks - The Brighter Side of News

‘Shape-shifting’ quantum computer uses light to switch between different tasksClavina combines linear and nonlinear quantum operations, advancing light-based quantum computing and error correction. Edited By: Shy Cohen/ICL Writer: Eleanor BarrandPublished Sep 7, 2026 4:07 PM PDTImperial researchers developed Clavina, a modular photonic quantum processor that expands what light-based quantum computers can do. (CREDIT: Shutterstock) Share this storyImperial College London researchers developed a modular photonic processor called Clavina that combines programmable linear optical circuits with nonlinear quantum modules in one reconfigurable architecture.The system demonstrated 100-mode Gaussian boson sampling, quantum correlations across 8,000 time bins, Schrödinger cat states and quasi-deterministic production of error-correction resources called GKP states at roughly 2,000 per second.Clavina is not yet a fault-tolerant quantum computer, but its ability to reuse and swap functional modules could help photonic machines move beyond specialized experiments toward more general-purpose quantum processors.Light may become one of the most powerful tools for future quantum computers. Photons, the tiny particles that make up light, can carry quantum information quickly and with relatively low noise. But they also create a stubborn problem: they do not naturally interact very strongly.That weak interaction has limited many light-based quantum machines. They can perform useful linear operations, but they struggle with the nonlinear steps needed for full quantum computing.Researchers from Imperial College London’s Department of Physics and external collaborators have now developed a new architecture called Clavina, which combines programmable linear optics with specialized nonlinear quantum modules in one system.A Processor Inspired By Modern ComputersTraditional photonic quantum systems often work like custom-built machines. They are designed for one task, and adapting them can require major hardware changes.A section of the Clavina quantum photonic processor. Fast electro-optic modulators switch the encoded time bins into functional modules and perform fast programming of the processor. (CREDIT: Imperial College London) Clavina takes a different approach. Its design draws inspiration from modern computer processors, which combine specialized parts for different jobs.A classical computer may use separate components for graphics, AI and general processing. In a similar spirit, Clavina routes quantum light between different functional modules.A central control unit directs information through a programmable optical network and nonlinear units. That means the same hardware can support different computations without a full redesign.Dr Shang Yu, lead author and Marie Skłodowska-Curie Fellow at Imperial, said the team wanted a major leap in function.“We set out to build a photonic quantum processor that provides a step change in functionality over our previous designs.Our new architecture ‘Clavina’ is scalable, modular and extensible, allowing different functional modules to be used depending on the computational task we wish to perform.”Why Light Needs Nonlinear HelpPhotons are attractive for quantum computing because they travel well through optical systems. They can also carry information in time, phase and other light-based properties.Team members of RUQu lab at Imperial College London, led by Raj Patel (Left to right: Shang Yu, Zhenghao Li.

Dayne Marcus Lopena, Raj Patel, Shana Winston, Ewan Mer, Yazeed Alwehaibi, Gerard Jiminez Machado). (CREDIT: Imperial College London) But photons usually pass through one another without interacting. That makes some quantum operations difficult to perform.Linear operations can guide, split, mix and measure light. They are useful, but not enough for universal quantum computation.Nonlinear operations allow quantum states to change in richer ways. They help create special states and interactions needed for advanced computing.Clavina brings these two worlds together. It keeps the flexibility of programmable optical circuits while adding nonlinear modules when tasks require them.A Modular Quantum MachineThe architecture uses time bins, which are short time slots carrying quantum information. Instead of building one physical path for every mode, the system reuses shared components.The control unit decides when light should pass through each part. It can route signals through linear networks, squeezing modules or nonlinear Kerr interaction units.This design reduces hardware demands and allows future expansion. New light sources, detectors or functional modules can be added without rebuilding everything.An extensible universal photonic quantum computing architecture and example applications. (CREDIT: Nature Photonics) Co-author Ying Dong of China Jiliang University said that flexibility marks a major shift.“Previously, researchers would have to build photonic hardware tailored to a specific task. The ability to switch in different functional modules in our architecture enables a single set of hardware to perform multiple functions without overhauling the design.Clavina can solve graph problems, simulate quantum systems or generate large entangled states and resource states for error correction.”Testing Large Optical CircuitsThe team first tested Clavina’s linear optical performance. The system performed large-scale transformations on optical modes with strong phase stability.That stability matters because light-based quantum circuits depend on precise interference. Small drifts can damage results.Researchers ran a 100-mode Gaussian boson sampling experiment. This tested whether the system could control many optical modes and handle adjustable squeezed-state inputs.They also generated scalable cluster states using time-multiplexed squeezed light pulses. Cluster states are important resources for some quantum computing approaches.Scalability and nonlinearity through modular functional units. (CREDIT: Nature Photonics) The system confirmed quantum correlations across 8,000 time bins. That showed it could support sustained multimode entanglement over a large optical structure.Creating Exotic Quantum StatesThe next challenge involved non-Gaussian states. These states are hard to make but essential for universal quantum computing.Clavina used a boosted photon-number-state generator to create nonlinear resource states. The system then routed them into an inline squeezing unit.This allowed the platform to produce Schrödinger cat states. In quantum physics, a cat state describes a superposition of distinct light states.The architecture also helped make larger cat states through a process called breeding. Pairs of smaller cat states interfered, producing states with larger amplitude.These unusual states are not just scientific curiosities. They can support advanced error correction, which future quantum computers will need.Engineering non-Gaussian resources for quasi-deterministic GKP state generation. (CREDIT: Nature Photonics) A Step Toward Error CorrectionOne major test involved Gottesman-Kitaev-Preskill states, known as GKP states. These states help encode quantum information so errors can be detected and corrected.Quantum computers are fragile. Noise, loss and imperfect operations can corrupt information. Without error correction, large quantum machines cannot run reliable calculations.Previous photonic methods produced GKP states only probabilistically. That means researchers could not make them on demand with high reliability.Clavina offered a more consistent route. The system used cat states, measurements and real-time feedback to produce GKP states at about 2,000 per second.Dr Raj Patel, UKRI Future Leaders Fellow and leader of Imperial’s photonic quantum computing programme, said the architecture supports key resources.“Our architecture allows the generation of exotic quantum states such as Schrödinger cat states and Gottesman-Kitaev-Preskill states, which are valuable resources for fault-tolerant quantum computing. The nonlinear operations in our architecture provide a universal gate set at the physical level which is required for the future implementation of bosonic error-correcting codes.”Nonlinear resource for simulating a Bose–Hubbard Hamiltonian. (CREDIT: Nature Photonics) Simulating Quantum MatterThe researchers also used Clavina to simulate the Bose-Hubbard model. This model describes how quantum particles move and interact across a lattice.It is important in condensed matter physics because it helps researchers study many-body quantum behavior. These systems become difficult for ordinary computers to model as they grow.Clavina handled the simulation by combining linear and nonlinear operations. The linear unit represented particles moving between sites. The nonlinear unit represented interactions when particles shared a site.Dr Jinzhao Sun of Queen Mary University of London said this versatility matters.“The versatility of our hardware enables us to integrate nonlinear operations and linear operations. We use these operations to perform simulations involving many-body interactions which are also restricted on superconducting quantum computers.”The system allowed researchers to tune interactions and observe how quantum states changed over time.What Still Needs ImprovementClavina is not yet a finished fault-tolerant quantum computer. The GKP states need higher quality, lower loss and stronger squeezing for practical error correction.The nonlinear operations also need further refinement. Better detectors, faster modulators and more efficient optical components could improve performance.The researchers also see a future path toward integrated photonic chips. Smaller chip-based systems could reduce optical loss and make the architecture more stable.Still, the study shows that photonic quantum machines can move beyond fixed, task-specific designs. A modular platform can adapt, expand and combine different quantum functions.That makes Clavina less like a single-purpose experiment and more like a framework for future processors.Dig deeper into photonic quantum computingThese resources examine modular optical processors, GKP states, cluster-state computing and the technologies needed to scale quantum computers based on light.A versatile single-photon-based quantum computing platform: Demonstrates a reconfigurable photonic processor capable of gate-based computation, boson sampling, quantum machine learning and molecular-energy calculations on shared hardware. (Nature Photonics, 2024)Scaling and networking a modular photonic quantum computer: Describes the Aurora system, which connects dozens of photonic chips and demonstrates a modular architecture for scaling optical quantum processors. (Nature, 2025)Integrated photonic source of Gottesman–Kitaev–Preskill qubits: Shows how GKP states, important resources for bosonic error correction, can be generated using an integrated low-loss photonic platform. (Nature, 2025)Resource-efficient photonic quantum computation with high-dimensional cluster states: Explores an alternative method for constructing photonic cluster states while reducing resources required for measurement-based quantum computation. (Nature Photonics, 2024)Recent progress towards large-scale integrated photonic quantum computation: Reviews photonic chips, sources, interference networks, detectors and the remaining engineering challenges on the path toward universal fault-tolerant optical computing. (npj Nanophotonics, 2026)Research findings are available online in the journal Nature Photonics.The original story "‘Shape-shifting’ quantum computer uses light to switch between different tasks" is published in The Brighter Side of News.Related StoriesWorld’s first superconducting quantum heat engine looks to transform quantum computingQuantum cat strategy reduces errors and speeds up quantum computingScientists unlock scalable entanglement for next-generation quantum computingLike these kind of feel good stories?

Get The Brighter Side of News' newsletter.Share this storyGlobal Good NewsInnovation NewsLightProcessorsQuantumquantum computersQuantum Computingquantum processorResearchScienceTechnologyShy CohenScience and Technology WriterShy CohenWriter Shy Cohen is a Washington-based science and technology writer covering advances in artificial intelligence, machine learning, and computer science. Having published articles on MSN, AOL News, and Yahoo News, Shy reports news and writes clear, plain-language explainers that examine how emerging technologies shape society. Drawing on decades of experience, including long tenures at Microsoft and work as an independent consultant, he brings an engineering-informed perspective to his reporting. His work focuses on translating complex research and fast-moving developments into accurate, engaging stories, with a methodical, reader-first approach to research, interviews, and verification.Latest StoriesInnovationsFirst-ever floating titanium looks to transform marine constructionJoseph ShavitSpaceFaint manganese signals could reveal how the universe built its elementsJoseph ShavitHealthBiology’s four-letter DNA alphabet just expanded to eightJoshua Shavit