Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

GNoME, the AI driver platform, created 41 new materials in 17 days alone, all of which are much faster and more accurate than human beings.

2024-04-13 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > IT Information >

Share

Shulou(Shulou.com)12/24 Report--

CTOnews.com December 4 news, according to "Science and Technology Daily" reported that "Nature" magazine published on November 30 research shows that the latest AI-driven platform GNoME (Materials Discovery Graphics Network) can discover and synthesize new inorganic compounds on its own, which has discovered more than 2.2 million stable structures and can create 41 new materials in 17 days alone, faster and more accurate than humans.

▲ The new material Ba 6Nb 7O21 calculated by GNoME, the source of the Berkeley Laboratory Materials Project, has this tool, and the number of feasible new materials created by known human techniques will multiply. GnoME discovered more than 2.2 million stable structures and improved the accuracy of structural stability prediction to more than 80%, with an accuracy of 33% per 100 trials when predicting components, compared to only 1% in previous work.

DeepMind team proposes a computational model that can improve the efficiency of material discovery through large-scale active learning. The program is able to train using existing literature and generate a wide variety of candidate structures for potential compounds, which are subsequently refined through a series of learning.

The DeepMind team announced in its press release that the discovery of 2.2 million new stable structures is equivalent to nearly 800 years of knowledge.

In addition, the UC Berkeley team has developed an automated laboratory system (A-Lab) that, trained on existing scientific documentation and combined with active learning, can create up to five initial synthetic recipes for proposed compounds. It can then also use the robotic arm to perform experiments to synthesize compounds in powder form.

If a recipe yields less than 50%, the system will adjust the recipe to continue the experiment, ending when the target is successfully achieved or all possible recipes are exhausted. In detail, 355 experiments were performed by the automated laboratory system during these 17 days, yielding 41 of the 58 proposed compounds. If this task were left to human researchers, it would take months of guesswork, experimentation and so on.

CTOnews.com with Nature related papers: published by DeepMind, published by Berkeley Lab

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

IT Information

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report