可持续神经形态技术
Search documents
长征十二号甲运载火箭首飞入轨成功,回收未取得预期成效;能自行修复的量子计算机问世丨智能制造日报
创业邦· 2025-12-24 03:25
Group 1 - A biodegradable and ultra-low power artificial synapse has been developed by scientists at Ulsan Institute of Science and Technology, made from natural materials like shells, beans, and plant fibers, capable of extending memory retention to nearly 100 minutes, paving the way for sustainable neuromorphic technology and addressing electronic waste issues [2] - A self-repairing neutral atom quantum computer has been demonstrated by the American company Atom Computing, overcoming the challenge of atomic loss during operation, which lays the foundation for sustainable quantum computers and offers new insights for larger-scale applications [2] - The Long March 12A rocket successfully completed its maiden flight on December 23, 2025, achieving its second-stage orbital goal, although the first-stage recovery did not meet expectations, providing valuable data and practical experience for future technology iterations [2] Group 2 - A launch failure occurred with the Korean Innospace's rocket in Brazil, which crashed shortly after takeoff, leading to a 30% drop in Innospace's stock price, highlighting the risks associated with space launch operations [2]
可降解超低功耗人工突触研制成功 将记忆保持时间延长至近百分钟
Ke Ji Ri Bao· 2025-12-23 01:00
Core Insights - Researchers at Ulsan National Institute of Science and Technology have developed a fully biodegradable artificial synapse made from natural materials, which can extend memory retention time to nearly 100 minutes, addressing issues related to electronic waste and paving the way for sustainable neuromorphic technology [1][2]. Group 1: Performance and Structure - The artificial synapse features a layered structure resembling a mini "sandwich," with two ionic active layers sandwiching an ionic bonding layer made from cellulose acetate derived from plant stems, along with materials sourced from shells and legumes [1]. - This device operates with low energy consumption, requiring only 0.85 femtojoules per signal transmission, significantly lower than the 2.4 to 24 femtojoules needed by natural brain synapses [1]. Group 2: Biodegradability and Environmental Impact - The ionic active layer and bonding layer of the device can naturally decompose in soil within approximately 16 days, using environmentally safe materials that do not leave harmful residues after decomposition [2]. - The development addresses multiple challenges in artificial synapses, including ultra-low power consumption, stability, durability, and biodegradability, marking a significant advancement in creating neuromorphic devices that can safely interact with the environment and eventually dissolve [2]. Group 3: Application Potential - A simple robotic system demonstrated the device's application potential, where it can amplify signals in response to heat, triggering a mechanical hand to withdraw from a hot object, simulating human reflex behavior [2]. - This suggests future possibilities for developing environmentally friendly robots that can learn, respond to environmental stimuli, and harmlessly disappear after completing their tasks [2].
可降解超低功耗人工突触研制成功将记忆保持时间延长至近百分钟
Xin Lang Cai Jing· 2025-12-22 23:28
Core Insights - A biodegradable artificial synapse has been developed by scientists at Ulsan Institute of Science and Technology, which can extend memory retention time to nearly 100 minutes, marking a significant advancement in sustainable neuromorphic technology [1][2] Group 1: Performance and Structure - The artificial synapse is made from natural, environmentally friendly materials such as shells, legumes, and plant fibers, and features a layered structure that mimics brain synapse functions [1] - The device operates with extremely low energy consumption, requiring only 0.85 femtojoules per signal transmission, significantly lower than the 2.4 to 24 femtojoules needed by natural brain synapses [1] Group 2: Biodegradability and Environmental Impact - The ion-active layers and the binding layer of the device can naturally decompose in soil within approximately 16 days, ensuring that no harmful substances are left behind after degradation [2] - This development addresses multiple challenges in creating neuromorphic devices that are low-power, stable, durable, and biodegradable, paving the way for environmentally safe interactions [2] Group 3: Application Potential - A simple robotic system has demonstrated the potential applications of the device, where it can amplify signals in response to heat, mimicking human reflex actions [2] - The research indicates the possibility of developing eco-friendly robots that can learn, respond to environmental stimuli, and safely decompose after completing their tasks [2]