{"id":34063,"date":"2025-04-26T13:03:09","date_gmt":"2025-04-26T05:03:09","guid":{"rendered":"https:\/\/www.1ai.net\/?p=34063"},"modified":"2025-04-26T13:03:09","modified_gmt":"2025-04-26T05:03:09","slug":"%e4%bb%8e%e5%88%86%e9%92%9f%e5%88%b0%e7%a7%92%e7%ba%a7%ef%bc%9ameta-space-llama-%e6%a8%a1%e5%9e%8b%e7%99%bb%e9%99%86%e5%9b%bd%e9%99%85%e7%a9%ba%e9%97%b4%e7%ab%99%ef%bc%8c%e5%88%b7%e6%96%b0%e5%a4%aa","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/34063.html","title":{"rendered":"From Minutes to Seconds: Meta Space Llama Model Lands on the International Space Station, Setting New AI Speed in Space"},"content":{"rendered":"<p>April 26 News.<a href=\"https:\/\/www.1ai.net\/en\/tag\/meta\" title=\"[View articles tagged with [Meta]]\" target=\"_blank\" >Meta<\/a> The company made an announcement yesterday (April 25) that it has joined hands with <a href=\"https:\/\/www.1ai.net\/en\/tag\/booz-allen\" title=\"_Other Organiser\" target=\"_blank\" >Booz Allen<\/a>It's already in the<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%9b%bd%e9%99%85%e7%a9%ba%e9%97%b4%e7%ab%99\" title=\"[Sees articles with [International Space Station] labels]\" target=\"_blank\" >International Space Station<\/a>(ISS)<strong>The U.S. National Laboratory deployed Space Llama, which is based on the open source Llama 3.2.<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%ae%9a%e5%88%b6ai%e6%a8%a1%e5%9e%8b\" title=\"[SEES ARTICLES WITH [CUSTOMED AI MODEL] LABELS]\" target=\"_blank\" >Customized AI models<\/a>.<\/strong><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-34064\" title=\"5253c1f3j00svb7c000bhd000p000dxp\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2025\/04\/5253c1f3j00svb7c000bhd000p000dxp.jpg\" alt=\"5253c1f3j00svb7c000bhd000p000dxp\" width=\"900\" height=\"501\" \/><\/p>\n<p>Because the full \"model weights\" of the Llama model are publicly available, researchers can install and run the model on stand-alone hardware without an Internet connection, avoiding the risk of transmitting data through Earth's servers while improving data security and responsiveness.<\/p>\n<p>Space Llama integrates a number of cutting-edge technologies, including Booz Allen's A2E2 (AI for Edge Environments) framework, Wise &amp; Tech's Spaceborne Computer-2, NVIDIA's accelerated compute technologies (CUDA, cuDNN, and cuBLAS), as well as Meta's Llama visual AI capabilities.<\/p>\n<p>The system runs on a compact, low-energy device similar to those used in satellites. The team revealed that reasoning times for some tasks plummeted from minutes to just over a second.<\/p>\n<p>In addition, the model supports generative and multimodal AI, generating text or images on cue and processing multiple types of data, including text, visual, and audio. Researchers have direct access to technical reference documents and operational instructions without the need for an Internet connection, which greatly enhances efficiency.<\/p>\n<p>This project continues Booz Allen's August 2024 presentation of a generative Large Language Model (LLM) at ISS and Meta's November 2024 announcement of plans to make an optimized version of the Llama model available to the U.S. government and private partners.<\/p>\n<p>Bill Vass, Booz Allen's chief technology officer, said space innovation has long been limited by dependence on Earth connectivity, and Space Llama brings tools directly to the edge of space to facilitate rapid repair and maintenance of ISS national laboratories.<\/p>","protected":false},"excerpt":{"rendered":"<p>April 26, 2011 - Meta announced yesterday (April 25) that, in collaboration with Booz Allen, it has deployed Space Llama, a customized AI model based on the open source Llama 3.2, to the U.S. National Laboratory on the International Space Station (ISS). Because the full \"model weights\" of the Llama model are publicly available, researchers can install and run the model on standalone hardware without an Internet connection, avoiding the risk of transmitting data through servers on Earth while improving data security and responsiveness. Space Llama integrates a number of cutting-edge technologies, including Booz Allen's A2E,\" 1AI said, citing a blog post.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[6453,297,6454,6452],"collection":[],"class_list":["post-34063","post","type-post","status-publish","format-standard","hentry","category-news","tag-booz-allen","tag-meta","tag-6454","tag-ai"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/34063","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=34063"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/34063\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=34063"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=34063"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=34063"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=34063"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}