Kipu Quantum Demonstrates Neutral-Atom Quantum Feature Extraction for Molecular Toxicity Classification on QuEra Hardware

quantumcomputingreport.com Oct 17, 2026

Kipu Quantum Demonstrates Neutral-Atom Quantum Feature Extraction for Molecular Toxicity Classification on QuEra Hardware

Curated by Shubham · TechCapsules · Oct 17, 2026

AI-summarised brief · reviewed before publication

Kipu Quantum reported experimental results using QuEra Computing’s 256‑qubit Aquila neutral‑atom processor to perform analog quantum feature extraction for molecular toxicity classification. The team arranged 200 neutral atoms in a 14 × 15 grid and applied smooth driving pulses for under 4 µs, generating roughly 1,000 quantum features per molecule from 500 readout shots. These features were tested on the UCI Molecular Toxicity dataset across 100 random splits with classical models such as XGBoost, CatBoost, Random Forest, SVM and Google’s TabFM. On imbalanced metrics, quantum‑derived features raised balanced accuracy from 0.55 to 0.70 and doubled the F1‑score from about 0.30 to 0.60, while an SVM achieved an AUC of ~0.81. The approach avoids deep gate‑based circuits, leveraging continuous Rydberg‑state dynamics accessed via the Kipu Quantum Hub.

💡 Why It Matters

· By extracting high‑dimensional quantum features that outperform classical descriptors on tough, imbalanced datasets, the work demonstrates a practical advantage of analog neutral‑atom processors for real‑world drug‑discovery tasks.

Read full article on quantumcomputingreport.com

University of Oxford Entangles Multi-Species Atom Network and Extends Memory Storage Beyond…

Physicists at the University of Oxford demonstrated a heterogeneous quantum network that entangles ions of two different species—strontium (88Sr⁺) and calcium (43Ca⁺)—across separate trapped‑ion modules. The team produced up to four‑qubit Greenberger‑Horne‑Zeilinger states and stored the distributed entanglement in calcium memory qubits for more than 10 seconds, a lifetime exceeding the generation time by over 100 times. By co‑trapping a strontium optical interface with a calcium memory qubit in microfabricated Paul traps, the researchers avoided decoherence from networking lasers and achieved a record entanglement lifetime, advancing distributed quantum computing and sensing.

💡 The breakthrough shows that mixed‑species trapping can overcome the decoherence trade‑offs that have limited quantum networks, paving the way for scalable, long‑lived entanglement essential for practical quantum communication and computation.

quantumcomputingreport.com · Oct 17

Kipu Quantum and NTT DATA Partner to Scale Enterprise Quantum Computing and…

Kipu Quantum has partnered with NTT DATA to scale enterprise quantum computing, industrial optimization, and quantum machine learning (QML) solutions. The deal will embed Kipu’s hardware‑tailored hybrid quantum algorithms into existing enterprise technology stacks, with NTT DATA distributing the Kipu Quantum Hub and delivering Kipu’s training curriculum worldwide. The collaboration builds on a successful deployment with Komatsu Peru, where Kipu’s QML algorithms improved predictive maintenance for heavy equipment. Initial commercial applications will target supply‑chain logistics, predictive maintenance, and financial risk modeling, beginning in Latin America before global rollout.

💡 The partnership bridges cutting‑edge quantum software with a major global consultancy, accelerating real‑world adoption of quantum‑enhanced analytics across critical industries.

quantumcomputingreport.com · Oct 17

MIT Bengaluru Inaugurates Q-HUB@MAHE and Deploys QuantrolOx’s Open-Architecture VIDYAQAR Quantum System

ManipAl Institute of Technology (MIT Bengaluru), part of the Manipal Academy of Higher Education, inaugurated the Quantum‑Hub@MAHE on 13 October 2026. The centre, opened by Dr. Sridhar Vembu of Zoho and Dr. Marcus Rommel of ConScience, marks the launch of a multi‑vendor Indo‑European hardware consortium linking QuantrolOx (Finland), Bluefors (Finland), Qblox (Netherlands) and ConScience (Sweden). At its core is QuantrolOx’s open‑architecture VIDYAQAR system, which runs on a 25‑qubit dilution refrigerator and enables modular testing, benchmarking and one‑click calibration of quantum components via a Python‑based SDK. Phase 1 of India’s National Quantum Mission roadmap aims to scale to 150‑qubit proof‑of‑concept platforms and eventually systems exceeding 1,000 qubits, while training 100 quantum engineers by December 2026.

💡 The hub gives Indian researchers immediate access to a vendor‑agnostic testbed, accelerating home‑grown hardware development and closing the skilled‑labor gap in the nation’s quantum agenda.

quantumcomputingreport.com · Oct 17

RIKEN Demonstrates High-Fidelity Simultaneous Control of Silicon Spin Qubits Above 99.9% Threshold

RIKEN and collaborators reported simultaneous single‑qubit gates on a five‑qubit 28Si/SiGe quantum‑dot array using one shared microwave line. By identifying off‑resonant AC‑Stark shifts as the dominant source of parallel‑drive crosstalk, they introduced an N(N‑1) pairwise calibration protocol that linearly compensates phase errors. The method preserves gate fidelities above 99.9 % for all five qubits, surpassing the fault‑tolerant threshold required for quantum error correction. Idle silicon spin qubits dephase in ~10 µs, while driven qubits reach T2,R ≈ 100 µs, making concurrent operations essential for cyclic error‑correction cycles. The team also suggests that integrating back‑end‑of‑line nanomagnets could double field gradients, enabling 2–3 MHz gates and scaling to ten‑qubit modules in commercial silicon processes. The calibration requires only measurement overhead and can be automated for arrays.

💡 Achieving >99.9 % simultaneous gate fidelity unlocks practical error‑corrected silicon processors, moving the platform from laboratory demos toward scalable manufacturing.

quantumcomputingreport.com · Oct 17

Chalmers Next Labs, ConScience, and NQFF Validate Superconducting Qubit Chip Stability Across…

Chalmers Next Labs, ConScience, and Singapore’s National Quantum Federated Foundry (NQFF) confirmed that a 15‑qubit superconducting processor fabricated in Gothenburg retains its performance when measured in both Sweden and Singapore. The joint study, part of the WACQT–NQFF Twin Testbed funded by Sweden’s Vinnovau under the QUASI framework, evaluated the QiB15 chip across two geographically distinct facilities. Results showed no degradation from shipping, environmental, or magnetic field variations, validating cross‑facility consistency. The collaboration supports a 2025 Memorandum of Understanding between Singapore’s National Quantum Office and the Quantum Sweden Innovation Platform and will present full data at APS 2027.

💡 Demonstrating that quantum hardware can be reliably benchmarked across borders removes a critical barrier for global supply chains, enabling manufacturers to certify components without repeated local testing. This paves the way for scalable, standardized quality assurance in the emerging quantum industry.

quantumcomputingreport.com · Oct 17

Open Quantum Launches Native Qrisp Integration for High-Level QPU Execution Across IonQ,…

Open Quantum, a joint effort between Quantum Rings and qBitTensor Labs, has released a native Qrisp integration that lets developers compile and run high‑level Qrisp programs directly on IonQ, Rigetti, IQM, and AQT quantum processors. The Python package “openquantum-sdk[qrisp]” eliminates the need for vendor‑specific gate wiring by translating Qrisp’s typed variables into Qiskit circuits, validating against each QPU’s gate set and coupling map, and submitting jobs via OpenQASM. This plug‑in supports both trapped‑ion and superconducting hardware, advancing European open‑source quantum software initiatives under Horizon Europe’s Q‑VERSE consortium.

💡 By removing low‑level qubit management, the integration speeds up algorithm prototyping across diverse hardware, enabling researchers to focus on problem logic rather than device constraints.