The convergence of Quantum with AI
Qubits built from superconducting circuits cooled to near absolute zero. The most industrially mature approach, favoured by IBM, Google, and Rigetti; fast gates, but short coherence times and demanding cryogenics.
61 stories
QpiAI has inaugurated an 8-inch quantum chip foundry in Bengaluru, targeting QPUs scalable to 10,000 qubits. The facility is intended to support in-house fabrication of superconducting quantum chips for QpiAI's quantum processors.
The foundry could allow QpiAI to iterate on superconducting qubit designs quickly enough to deliver a 100-qubit processor within two years.
Researchers at the Universidade Federal de São Carlos have demonstrated a pulse-engineering technique that realizes arbitrary single-qubit rotations with a single shaped control pulse. The method, reported by Quantum Zeitgeist, removes the need for multi-pulse composite sequences and was validated on superconducting qubit hardware.
Superconducting quantum computing platforms could integrate this single-pulse gate scheme to shorten single-qubit gate times and reduce error accumulation in near-term processors.
An arXiv preprint presents an analytical blueprint for implementing X-gates at 99.999% fidelity on existing superconducting qubits using strong driving. The approach derives pulse shapes analytically rather than via numerical optimization and is claimed to be compatible with current transmon hardware parameters.
If the analytical pulses are experimentally validated, superconducting processors could achieve five-nines single-qubit gate fidelity through software-level control changes alone, improving baseline error rates for near-term error-correction experiments.
A preprint on arXiv proposes a scheme for homomorphic aggregation of continuous-variable Gottesman-Kitaev-Preskill (GKP) states. The work describes combining multiple GKP-encoded qubits while preserving error-correction structure, without full decoding of the logical information. It addresses operations on bosonic codes for fault-tolerant quantum computing.
This could allow near-term experimental platforms using GKP states, such as superconducting cavity QED or photonic systems, to test distributed or multi-qubit operations with reduced decoding overhead.
A superconducting device has been reported that implements continuous-variable quantum computing, an approach usually associated with photonic systems. The work was published by Quantum Zeitgeist and appears to originate from a Brazilian federal university.
If the superconducting continuous-variable device can be integrated with existing circuit-QED control electronics, it could enable hybrid discrete-variable/continuous-variable processors on the same chip within two years.
Researchers have demonstrated a chip-scale isolator for quantum systems that suppresses back-reflections by 30 dB, as reported by Quantum Zeitgeist. The device targets cryogenic microwave signal chains, where reflected signals can disturb qubit operation.
This could allow near-term superconducting quantum processors to replace bulky off-chip circulators with integrated isolators, reducing thermal load and wiring complexity.
An arXiv preprint proposes a scheme for holonomic quantum gates driven by continuous measurement in bosonic error-correcting codes, specifically GKP and cat states. The work is theoretical and develops geometric gate constructions that could be robust to certain control errors. No experimental demonstration is reported.
The proposal could be translated into an experimental demonstration of continuous-measurement-driven holonomic gates on superconducting cavity GKP or cat qubits within two years.
Researchers demonstrated dynamic quantum circuits on a hybrid superconducting qubit-cavity processor, showing computational advantages over static circuits, as reported on arXiv.
The demonstrated dynamic circuit techniques could be integrated into near-term error mitigation protocols, improving the effective fidelity of noisy superconducting processors.
D-Wave has developed a new qubit gate that reduces error rates by a factor of ten, as reported by Quantum Zeitgeist. This improvement is part of their superconducting gate-model quantum computing program. The gate demonstrates significantly enhanced fidelity compared to previous implementations.
This could enable D-Wave to demonstrate a logical qubit with lower error rates within the next two years, provided they can integrate the gate into a multi-qubit system with adequate connectivity.
D-Wave published a paper in Nature detailing an advance in dual-rail quantum computing. The technique encodes a logical qubit in two physical superconducting qubits to reduce errors and extend coherence times.
D-Wave could integrate dual-rail encoded qubits into its next-generation quantum processors within two years, potentially offering gate-model operations with lower error rates.
D-Wave published research detailing a dual-rail qubit gate operation on its superconducting processor, aimed at enabling quantum error correction. The dual-rail encoding uses two physical qubits to represent a logical qubit, allowing for error detection. This work represents a move beyond annealing into gate-based fault-tolerant quantum computing.
D-Wave could demonstrate a logical qubit with error detection within two years, integrating the dual-rail gate into a small-scale device.
A paper on arXiv proposes a bias-preserving CNOT gate for cat qubits realized via a vacuum-conditional beam-splitter. The gate is designed to preserve the dominant error bias of cat qubits, preventing conversion of bit-flip to phase-flip errors during two-qubit operations. This addresses a key challenge in fault-tolerant quantum computing with biased-noise bosonic codes.
Experimental demonstration of this gate could enable low-overhead fault-tolerant operations in cat-qubit processors, bringing such architectures closer to error-corrected operation within two years.
IQM Quantum Computers has shipped its first quantum computer to a US customer, Oak Ridge National Laboratory (ORNL), marking the Finnish startup's entry into the American market. The company also disclosed revenue for the first time, signaling commercial progress beyond research contracts.
ORNL's adoption could pave the way for additional US government and enterprise orders for IQM, particularly as the lab validates performance and publishes results, building credibility for the technology.
A new preprint proposes a fault-tolerant quantum computing architecture based on a microwave Cat Bus, using cat states for robust qubit communication and error correction.
If the Cat Bus can be implemented with existing superconducting technology, it could enable a demonstration of a logical qubit with reduced overhead within two years.
Researchers have demonstrated a neural network approach to inversely design transmon qubit geometries based on desired electromagnetic properties. The method generates component-level designs from target frequencies and anharmonicities, bypassing iterative simulation.
If the inverse design approach proves robust for real fabrication tolerances, it could enable rapid prototyping of novel qubit designs with tailored properties, reducing the design cycle from weeks to hours.
IQM Quantum Computers and Deutsche Bahn have successfully executed a hybrid quantum algorithm for railway scheduling. The collaboration applied quantum computing to optimize train timetables, demonstrating a practical use case for the technology.
Deutsche Bahn could expand the hybrid quantum scheduling approach to a larger subset of its network, moving from proof-of-concept to a limited operational pilot, provided IQM's hardware scales to handle larger problem instances within the next two years.
Researchers used a quantum computer to perform a computational task that is beyond the practical reach of classical supercomputers, with verification confirming the correctness of the result.
If the verification is robust, this result could shift investor and industry perception from quantum computing as a long-term play to a near-term practical tool, leading to a surge in funding for applied quantum computing startups and industrial consortia.
Cleveland Clinic and IBM announced the development of a quantum machine learning model designed to predict cancer neoantigens, potentially improving the selection of immunogenic peptide sequences for personalized cancer vaccines.
The model is refined on larger datasets and integrated into a hybrid classical–quantum pipeline for neoantigen screening in early‑phase clinical trials.
BlueQubit, a quantum software startup, provided support for a research study claiming quantum advantage using error mitigation techniques. The study reportedly demonstrated a computational task where a noisy quantum processor, aided by error mitigation, outperformed classical computers.
If error mitigation techniques can be reliably scaled to slightly larger circuits, this could enable practical quantum advantage for niche problems in optimization or simulation within two years, before full fault tolerance is achieved.
IBM and Qedma demonstrated quantum advantage in simulating the Floquet Ising model on a superconducting quantum processor, using Qedma's error mitigation to extract accurate dynamics beyond classical verification.
This could catalyze adoption of quantum simulation for short-time dynamics in materials science, as error mitigation proves sufficient to extract physically meaningful results on near-term devices.