News Summaries

New Chip Design Boosts Quantum Computer Speed and Accuracy

A team of physicists at the Chinese Academy of Sciences has tackled a major hurdle in building practical quantum computers. Quantum bits, or qubits, need to stay perfectly quiet to preserve their delicate quantum state, yet they must also be read quickly and reset for the next calculation—a tricky balance that has limited progress. The researchers created a superconducting chip that embeds tiny, adjustable filters. These filters can be tuned on‑the‑fly to sharpen the signal when a qubit is measured, achieving a record‑high readout accuracy of 99.3% without the need for extra ultra‑low‑noise amplifiers. At the same time, the filters open a fast “drain” that wipes the qubit clean in just 75 to 200 nanoseconds, far quicker than previous methods and with error rates under 1%. The prototype chip houses 24 qubits, 38 couplers and eight of these tunable filters, showing the design can scale up. By keeping qubits coherent while allowing rapid, reliable readout and reset, this breakthrough paves the way for quantum error‑correction and more complex algorithms, bringing fault‑tolerant quantum computers a step closer to reality.

AI Detects Dozens of Hidden Worlds in NASA’s TESS Data

Astronomers at the University of Warwick have harnessed a new artificial‑intelligence pipeline, dubbed RAVEN, to sift through more than 2.2 million stars observed by NASA’s Transiting Exoplanet Survey Satellite (TESS). The AI‑driven search has validated over 100 exoplanets – including 31 brand‑new discoveries – and pinpointed thousands of promising candidates. By concentrating on planets that zip around their stars in less than 16 days, the team has produced the most precise estimate yet of how common these short‑period worlds are around Sun‑like stars. “RAVEN lets us analyze massive data sets consistently and objectively,” said Dr. David Armstrong, associate professor at Warwick. “Because the pipeline is rigorously tested, this isn’t just a wish‑list of possible planets; it’s a reliable sample we can use to map planetary demographics.” The researchers have released interactive tools and catalogs so other scientists can explore the findings and select targets for follow‑up with ground‑based telescopes and upcoming missions such as ESA’s PLATO. The work appears in the Monthly Notices of the Royal Astronomical Society and promises to accelerate the hunt for new worlds.

Agile Robots Teams Up with Google DeepMind to Power Smarter Factories

Munich‑based robotics startup Agile Robots announced a new strategic research partnership with Google DeepMind on Tuesday. The deal will see Agile Robots embed DeepMind’s Gemini Robotics foundation models into its next‑generation machines, while the data gathered by those bots will help fine‑tune Gemini’s underlying AI. Together, the two companies plan to test, refine, and roll out autonomous robots across a range of industrial settings, from electronics assembly lines and car factories to data‑center maintenance and warehouse logistics. By combining Agile Robots’ expertise in hardware design and real‑world deployment with DeepMind’s cutting‑edge machine‑learning research, the partnership aims to create bots that can make decisions on the fly, adapt to unexpected changes, and operate with minimal human oversight. The collaboration reflects a broader trend of AI labs joining forces with robotics firms to accelerate the commercialization of intelligent automation. Earlier this month, German startup Neura Robotics partnered with Qualcomm to use its new IQ10 processor for mobile and humanoid robots. As the demand for flexible, self‑learning machines grows, such alliances are becoming a fast‑track way to bring sophisticated, AI‑driven robots from the lab to the factory floor.