Using a large language model to search quantum circuits, researchers reduced both Toffoli gate counts and logical qubit counts in seven of eight component circuits.
FAU compared methods for predicting lung cancer from limited symptom and risk‑factor data; a D‑Wave hardware method achieved 91.90% accuracy.
They published resource estimates for Shor’s algorithm using cat qubits and LDPC‑cat codes, estimating RSA‑2048 could be broken in 3.14 days.
Pasqal’s neutral‑atom quantum processor observed much slower relaxation than theoretical predictions in a two‑dimensional quantum magnet. The results are published as a preprint.
IBM connected two superconducting quantum processors with four l-couplers, demonstrating an average 99% quantum state-transfer fidelity.
QunaSys simulated a quantum phase estimation circuit on a large-scale GPU cluster and achieved chemical accuracy for a four‑hydrogen model.
OptQC has published an arXiv paper proposing a photonic quantum-computing architecture that separates the offline resource-generation unit for error correction from the online computation unit.
They developed a method to iteratively correct correlated errors arising in cross-logical-qubit operations for trapped-ion systems, and compared logical error rates in large-scale simulations.
Quantinuum recorded five logical errors per one million operations on its trapped‑ion quantum computer Helios using the color code.
Google Quantum AI demonstrated magic-state cultivation on the Willow quantum processor, achieving 99.99% logical fidelity for T states.
Diraq says quantum computers should be evaluated not by qubit count but by logical-operation throughput and cost per operation.
Achieved entanglement generation rates exceeding 1 kHz between trapped-ion qubits and silicon-vacancy (SiV) quantum memory via a photonic interconnect.
Alice & Bob and ENS Lyon demonstrated a method to control two-photon dissipation in superconducting cat qubits using a tiny DC voltage instead of microwave drives.
A method that approximates a quantum state as a combination of basis functions and reconstructs it with fewer measurements than conventional methods was demonstrated in numerical simulations.
They are examining algorithms and implementation methods, using problems such as the one‑dimensional advection equation, to apply gate‑model quantum computers to automotive CAE and CFD.
IonQ applied a quantum circuit Born machine to change detection in high-resolution SAR imagery and reported results that outperform classical methods on non-Gaussian data.
They built a quantum–classical hybrid workflow to search operational configurations of a natural-gas network and ran a reduced-size problem on IonQ Forte‑1.
IonQ evaluated real-time error decoding on a single CPU using circuits simulating up to 408 logical qubits and more than 31.5 million quantum operations. The reported increase in execution time was as little as 0.02%.
IonQ and Synopsys incorporated quantum algorithms into pre-processing for engineering simulations, reducing end-to-end runtimes by 5.9–14.6%.
QuEra’s 2026 Quantum Readiness Survey found 45% of respondents prioritize a roadmap to fault tolerance. However, there was no clear consensus on the most promising quantum computing architecture.
IonQ, Oak Ridge National Laboratory (ORNL), NVIDIA and the University of Tennessee validated DQAOA-GPT, a method that uses generative AI to directly generate quantum optimization circuits. In GPU simulations, it suppressed the increase in circuit search time as subproblem size grew.
Alice & Bob and Hyperion Research recommended latency testing, software standardization, and joint workflow design to integrate HPC, quantum, and AI.
A quantum–HPC coupled simulation of a 12,635‑atom protein system and automation of the computational workflow were recognized.
IonQ has published a design and resource estimate for solving the secp256k1 elliptic-curve discrete logarithm problem on a fault‑tolerant quantum computer, estimating 19,397 physical qubits and 25.7 days per run.
They announced a method combining machine-learning potentials and orbital-free density functional theory to compute ion migration and local charge changes under electric fields.
A paper has been published demonstrating wiring‑reduced voltage control using time‑division multiplexing in a cryogenic environment for scaling trapped‑ion quantum processors.
Reports reducing physical qubit counts compared with the surface code at tested code sizes, while streamlining quantum logic operations and error correction.
QOBLIB, which compares quantum and classical optimization methods under common conditions, has been expanded through a published paper and the launch of a dedicated website.
Q-CTRL reports executing the quantum Fourier transform on up to 100 qubits on IBM Heron r3, extracting the correct frequency as the most frequent measurement outcome.
D-Wave demonstrated a two-qubit entangling gate with about 99.9% fidelity and roughly 500 ns gate time. The results were published in Nature.