A team of Chinese researchers has announced a major breakthrough for quantum computers: they built a “coherent quantum routing system” that works like an express lane for data inside a quantum processor. The new design uses extra energy levels in superconducting qubits to turn complicated routing steps into a handful of simple two‑qubit operations, dramatically shortening the circuit and cutting error rates. In tests on the “Origin Wukong” superconducting quantum computer, a single quantum router moved information with 98% efficiency, while a two‑layer routing network kept overall efficiency at 93%. Fidelity – a measure of how accurately quantum information is preserved – reached 94.8% for a single router and 82.4% for the two‑layer network. These numbers show that quantum data can be steered reliably using quantum addresses, a key requirement for scaling up Quantum Random Access Memory (QRAM). QRAM is the “data center” of a quantum computer, essential for algorithms like quantum search and machine‑learning. By solving the long‑standing bottleneck of QRAM scalability, the work moves China’s quantum research from isolated performance tweaks to real‑world, complex functions on actual hardware, promising cheaper, faster quantum calculations in the near future.
Read moreChina’s newest manufacturing park for high‑end medical devices shows how 5G combined with the industrial internet can turn a factory into a smart, self‑optimising hub. By linking thousands of sensors, RFID tags and low‑power devices to a 5G network, the park creates a real‑time digital twin of every machine and process. AI‑driven video cameras watch 127 4K feeds, spotting unsafe behavior, crowding or missing safety gear with 98.6 % accuracy, and instantly trigger alarms through phones, LED screens and public‑address systems. An energy‑management system reads electricity, water, gas and steam meters, predicts demand with LSTM models and uses reinforcement‑learning to cut overall consumption by 12.7 %, lowering carbon emissions by 9.3 %. The central platform lets engineers build low‑code apps for spare‑part recommendations, process‑parameter tuning and defect analysis, shifting production from “feel‑good” decisions to data‑driven ones. Edge‑computing nodes placed in the factory enable millisecond‑level closed‑loop control for climate, robotics and safety systems, while network slicing guarantees the ultra‑reliable links needed for critical tasks. The result is a factory that can see, analyse, decide, act and evolve on its own – a glimpse of the future of Chinese manufacturing powered by 5G.
Read moreChina is reshaping its space industry by moving satellite production from handcrafted, one‑off projects to fast, factory‑style assembly lines. Companies such as Galaxy Space, Hongqing Technology and Micro‑Nano Space are installing robotic workstations, automated testing rigs and digital design tools that can roll out a satellite in roughly 30 hours—far quicker than the months‑long cycles of the past. The goal is to cut costs, boost reliability and launch dozens of satellites each year for a new generation of broadband constellations. At the same time, China’s rocket sector is becoming more competitive and reusable, driving launch prices down and making it cheaper to put whole constellations into orbit. New satellite designs integrate 5G base stations, on‑board AI and laser links, turning them into space‑borne data hubs that can process information directly above the Earth. Experts say this will enable a seamless “air‑space‑ground” network, where satellites not only relay data but also store, compute and make decisions in real time. The broader vision includes a full‑service orbital economy—refueling stations, maintenance “4S shops” and space‑bus transfer vehicles—so that satellite internet becomes a reliable, everyday utility rather than a costly novelty. By standardising components, embracing modular platforms and pushing for international markets, China aims to turn its satellite internet ambitions into a robust, globally competitive industry.
Read moreA French quantum‑startup, Pasqal, has unveiled an AI “vibe‑coding” agent that can turn everyday English instructions into quantum‑computing code and then run that code on a real quantum processor. The breakthrough promises to shrink the steep learning curve that normally requires teams of specialists in materials physics and quantum algorithms. In trials, the agent was fed research papers describing complex physical phenomena and asked to produce and execute simulations that mimic those effects on a quantum computer. While the AI handled most of the heavy lifting, researchers still had to steer it, ensuring the generated programs obeyed the laws of physics—human expertise remains essential for experimental accuracy. Other groups are racing ahead, too. At NYU Abu Dhabi, physicist Nuhayla Inan and colleagues adapted the open‑source LLaMA language model to draft quantum code, slashing a four‑day coding effort down to a single day, even though the model can’t yet run the code autonomously. Together, these efforts show that large language models are becoming practical assistants for quantum research, lowering technical barriers and opening the field to a broader community of scientists.
Read more