Q-PLANET Begins Development of European Pilot Line for Neutral Atom Quantum Chips

Israeli Quantum Startups Eye Wall Street as SPAC Interest Accelerates

Israeli quantum computing companies Quantum Art and Classiq are in advanced talks to go public on Wall Street via SPAC mergers, with potential valuations between $2 billion and $5 billion each. Quantum Art develops trapped-ion quantum hardware and has raised ~$200 million, while Classiq builds cross-platform quantum software with partnerships including Nvidia, Microsoft, and AWS, and has raised a similar amount. The deals reflect a broader wave of quantum companies entering public markets — five have already gone public via SPACs in 2026, with another five in advanced talks, totaling ~$70 billion in combined valuation. Around 30 SPACs are reportedly searching for quantum acquisition targets. Israel has emerged as a significant quantum hub, ranking fifth globally in investment with ~$800 million deployed. The sector still faces major risks: most companies are pre-revenue, valuations rely on future expectations, and fault-tolerant quantum computers remain unachieved.

AI money watch: five funding rounds that matter today

Trong 24 giờ qua, các thương vụ gây quỹ đáng chú ý trong lĩnh vực AI và deep-tech bao gồm: Oratomic huy động 300 triệu USD Series A cho máy tính lượng tử nguyên tử trung hòa, Prime Intellect gọi vốn 130 triệu USD với định giá 1 tỷ USD cho công cụ AI tác tử doanh nghiệp, Gradium (Paris) mở rộng vòng seed vượt 100 triệu USD nhờ sự hậu thuẫn của Nvidia, Axle Energy (London) huy động 21 triệu euro cho cân bằng lưới điện thông qua thiết bị EV và năng lượng gia đình, cùng Bizay (Lisbon) đóng vòng Series D trị giá 55 triệu USD. Nvidia tham gia hai thương vụ, cho thấy chiến lược đầu tư vào các startup xây dựng trên phần cứng của mình, trong khi các nhà đầu tư ưu tiên mảng lượng tử, AI tác tử và cơ sở hạ tầng năng lượng/voice tại châu Âu hơn là các ứng dụng chatbot thông thường.

Lập trình viên nên theo dõi để hiểu cách Nvidia và các startup AI đang định hình tương lai của hardware-software tích hợp, từ kiến trúc lượng tử đến các công cụ agent AI chuyên dụng, giúp bạn cập nhật xu hướng công nghệ mới và ứng dụng thực tế trong phát triển phần mềm.

Nord Quantique Advances Quantum Error Correction with Sub-0.1% SPAM Errors

Nord Quantique has published research demonstrating quantum error correction of a single-mode grid state qubit with state preparation and measurement (SPAM) errors below 0.1% — roughly a 100-fold improvement over prior GKP-based systems, now on par with leading superconducting transmon platforms. The achievement uses a repeat-until-success post-selected stabilization protocol that improves state preparation fidelity without requiring complex real-time classical control. The protocol also supports high-fidelity magic state preparation, a resource-intensive requirement for universal quantum computation. The company's CEO cited this as a step toward fault-tolerant quantum computing by 2030, leveraging a 1:1 physical-to-logical qubit approach.

Harvest Now, Decrypt Later: The Threat to Key Exchange – Quantum Series, Part 2

Quantum computers pose a delayed but urgent threat to classical key exchange via the 'harvest now, decrypt later' attack — adversaries record encrypted traffic today and decrypt it once quantum hardware matures. Diffie-Hellman (including X25519/ECDH) is vulnerable to Shor's algorithm, making any long-lived sensitive data transmitted today at risk. ML-KEM (FIPS 203), NIST's standardized post-quantum Key Encapsulation Mechanism, is introduced as the quantum-safe alternative. Unlike DH's symmetric exchange, ML-KEM uses an asymmetric encapsulate/decapsulate model. ML-KEM-768 is recommended as the practical sweet spot. The post previews a hybrid approach combining both algorithms, to be detailed in Part 3.

Study Finds Imaginary Numbers May Not Be Essential to Quantum Mechanics

Physicists at Heinrich Heine University Düsseldorf and the German Aerospace Center have shown that quantum mechanics can be fully formulated using only real numbers, without requiring imaginary (complex) numbers. The team, led by Prof. Dagmar Bruß, identified that a postulate used in a landmark 2021 study (which concluded complex numbers are indispensable) was overly restrictive. By replacing it with a physically motivated alternative for system composition, they derived a class of theories that use only real numbers yet produce predictions experimentally indistinguishable from standard quantum mechanics. The findings were published in Physical Review Letters and highlighted by the American Physical Society.

Your Encryption Has an Expiration Date: France Turns Quantum Risk Into Procurement Law

France's cybersecurity agency ANSSI will stop certifying security products lacking quantum-resistant encryption starting 2027, with a full transition expected by 2030. This moves post-quantum cryptography from theoretical concern to a hard procurement requirement. The core risk is 'harvest now, decrypt later' — adversaries collecting encrypted data today to decrypt once quantum computers mature. Most organizations lack cryptographic inventories and don't know which systems rely on quantum-vulnerable algorithms like RSA or ECC. The US is pursuing a parallel path via NIST FIPS 203/204/205 standards, OMB inventory mandates, NSA's CNSA 2.0, and CISA guidance, but without France's bright-line certification cutoff. Practical steps include building a real cryptographic inventory, classifying data by secrecy lifetime, interrogating vendors on PQC readiness and crypto-agility, updating contracts with algorithm transition obligations, and engineering — not just declaring — the migration.

Quantum Companies Help Develop Hybrid AI for Immune-Targeting Peptides

Researchers from the Technical University of Denmark, ORCA Computing, and Sparrow Quantum developed a hybrid quantum-classical AI system for designing immune-targeting peptides. A photonic quantum processor using Gaussian boson sampling replaced the standard random noise inputs in a generative adversarial network, producing more structured starting patterns. The quantum-enhanced approach generated more predicted strong-binding peptides than conventional methods, with the largest gains for rare HLA variants that have limited training data. Laboratory tests confirmed that synthesized peptides formed stable complexes with their target immune proteins. The study does not claim quantum advantage — the quantum processor was small enough to simulate classically — but suggests quantum-generated probability distributions could serve as a useful supporting component in AI-driven vaccine and immunotherapy design pipelines.