{"id":2852,"date":"2023-01-24T14:50:38","date_gmt":"2023-01-24T14:50:38","guid":{"rendered":"https:\/\/salarydistribution.com\/machine-learning\/2023\/01\/24\/supersizing-ai-sweden-turbocharges-its-innovation-engine\/"},"modified":"2023-01-24T14:50:38","modified_gmt":"2023-01-24T14:50:38","slug":"supersizing-ai-sweden-turbocharges-its-innovation-engine","status":"publish","type":"post","link":"https:\/\/salarydistribution.com\/machine-learning\/2023\/01\/24\/supersizing-ai-sweden-turbocharges-its-innovation-engine\/","title":{"rendered":"Supersizing AI: Sweden Turbocharges Its Innovation Engine"},"content":{"rendered":"<div data-url=\"https:\/\/blogs.nvidia.com\/blog\/2023\/01\/24\/ai-sweden-berzelius\/\" data-title=\"Supersizing AI: Sweden Turbocharges Its Innovation Engine\" data-hashtags=\"\">\n<p>Sweden is outfitting its AI supercomputer for a journey to the cutting edge of machine learning, robotics and healthcare.<\/p>\n<p>It couldn\u2019t ask for a better guide than Anders Ynnerman (above). His signature blue suit, black spectacles and gentle voice act as calm camouflage for a pioneering spirit.<\/p>\n<p>Early on, he showed a deep interest in space, but his career took a different direction. He established the country\u2019s first network of supercomputing centers and went on to pioneer scientific visualization technologies used in hospitals and museums around the world.<\/p>\n<p>Today, he leads Sweden\u2019s largest research effort, <a href=\"https:\/\/wasp-sweden.org\/\">WASP<\/a> \u2014 the Wallenberg Artificial Intelligence, Autonomous Systems and Software Program \u2014 focused on AI innovation.<\/p>\n<h2><b>The Big Picture<\/b><\/h2>\n<p>\u201cThis is a year when people are turning their focus to sustainability challenges we face as a planet,\u201d said the Link\u00f6ping University professor. \u201cWithout advances in AI and other innovations, we won\u2019t have a sustainable future.\u201d<\/p>\n<p>To supercharge environmental efforts and more, Sweden will upgrade its <a href=\"https:\/\/blogs.nvidia.com\/blog\/2021\/03\/23\/ai-supercomputer-sweden\/\">Berzelius supercomputer<\/a>. Based on the <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/dgx-superpod\/\">NVIDIA DGX SuperPOD<\/a>, it will deliver nearly half an exaflop of AI performance, placing it among the world\u2019s 100 fastest AI supercomputers.<\/p>\n<p>\u201cA machine like Berzelius is fundamental not only for the results it delivers, but the way it catalyzes expertise in Sweden,\u201d he said. \u201cWe\u2019re a knowledge-driven nation, so our researchers and companies need access to the latest technology to compete.\u201d<\/p>\n<h2><b>AI Learns Swedish<\/b><\/h2>\n<p>In June, the system <a href=\"https:\/\/blogs.nvidia.com\/blog\/2022\/06\/19\/ai-sweden-nlp\/\">trained GPT-SW3<\/a>, a family of large language models capable of drafting a speech or answering questions in Swedish.<\/p>\n<p>Today, a more powerful version sports 20 billion parameters, a popular measure of a neural network\u2019s smarts. It can help developers write software and handle other complex tasks.<\/p>\n<p>Long term, researchers aim to train a version with a whopping 175 billion parameters that\u2019s also fluent in Nordic languages like Danish and Norwegian.<\/p>\n<p>One of Sweden\u2019s largest banks is already exploring use of the latest GPT-SW3 variant for a chatbot and other applications.<\/p>\n<h2><b>A Memory Boost<\/b><\/h2>\n<p>To build big AIs, Berzelius will add 34 <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/dgx-a100\/\">NVIDIA DGX A100 systems<\/a> to its cluster of 60 that makeup the SuperPOD. The new units will sport GPUs with 80GB of memory each.<\/p>\n<figure id=\"attachment_62051\" aria-describedby=\"caption-attachment-62051\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/blogs.nvidia.com\/wp-content\/uploads\/2023\/01\/Berzelius-and-Ynnerman-scaled.jpg\"><img decoding=\"async\" loading=\"lazy\" class=\"size-large wp-image-62051\" src=\"https:\/\/blogs.nvidia.com\/wp-content\/uploads\/2023\/01\/Berzelius-and-Ynnerman-672x356.jpg\" alt=\"Anders Ynnerman with Sweden's Berzelius AI supercomputer\" width=\"672\" height=\"356\"><\/a><figcaption id=\"caption-attachment-62051\" class=\"wp-caption-text\">Ynnerman with Berzelius at the system\u2019s March 2021 launch.<\/figcaption><\/figure>\n<p>\u201cHaving really fat nodes with large memory is important for some of these models,\u201d Ynnerman said. Atos, the system integrator, is providing \u201ca very smooth ride getting the whole process set up,\u201d he added.<\/p>\n<h2><b>Seeking a Cure for Cancer<\/b><\/h2>\n<p>In healthcare, a <a href=\"https:\/\/www.scilifelab.se\/data-driven\/\">data-driven life sciences program<\/a>, funded by the Wallenberg Foundation, will be a major Berzelius user. The program spans 10 universities and will, among other applications, employ AI to understand protein folding, fundamentally important to understanding diseases like cancer.<\/p>\n<p>Others will use Berzelius to improve detection of cancer cells and navigate the massive mounds of data in human genomes.<\/p>\n<p>Some researchers are exploring tools such as <a href=\"https:\/\/developer.nvidia.com\/omniverse-platform\/ace\">NVIDIA Omniverse Avatar Cloud Engine<\/a> and <a href=\"https:\/\/developer.nvidia.com\/bot-maker-early-access\">NVIDIA BotMaker<\/a> to create animated patients. Powered by GPT-SW3, they could help doctors practice telemedicine skills.<\/p>\n<h2><b>Robots in Zero Gravity<\/b><\/h2>\n<p>Sweden\u2019s work in image and video recognition will get a boost from Berzelius. Such algorithms advance work on the autonomous systems used in modern factories and warehouses.<\/p>\n<p>One project is exploring how autonomous systems act in space and undersea. It\u2019s a topic close to the heart of a recent addition to WASP, researcher Christer Fuglesang, who was named Sweden\u2019s first astronaut in 1992.<\/p>\n<p>Fuglesang went to the International Space Station in 2006 and 2008. Later, as a professor of physics at Sweden\u2019s Royal Institute of Technology, he collaborated with Ynnerman on live shows about life in space, presented in the <a href=\"https:\/\/wisdomeproject.se\/english\/\">WISDOME dome theater<\/a> at the <a href=\"https:\/\/visualiseringscenter.se\/en\">Visualization Center C<\/a> Ynnerman founded and directs.<\/p>\n<p>Thanks to his expertise in visualization, \u201cI can go to Mars whenever I want,\u201d Ynnerman quipped.<\/p>\n<p>He took NVIDIA founder and CEO Jensen Huang and Marcus Wallenberg \u2014 scion of Sweden\u2019s leading industrial family \u2014 on a tour of outer space at the dome to mark the Berzelius upgrade. The dome can show the Martian surface in 8K resolution at 120 frames per second, thanks to its use of 12 <a href=\"https:\/\/resources.nvidia.com\/en-us-general-rtx\/quadro-rtx-8000-data-sheet\">NVIDIA Quadro RTX 8000 GPUs<\/a>.<\/p>\n<h2><b>Inspiring the Next Generation<\/b><\/h2>\n<p>Ynnerman\u2019s algorithms have touched millions who\u2019ve seen visualizations of Egyptian mummies at the British Museum.<\/p>\n<p>\u201cThat makes me even more proud than some of my research papers because many are young people we can inspire with a love for science and technology,\u201d he said.<\/p>\n<p>A passion for science and technology has attracted more than 400 active Ph.D. candidates so far to WASP, which is on the way to exceeding its goal of 600 grads by 2031.<\/p>\n<p>But even a visualization specialist can\u2019t be everywhere. So Ynnerman\u2019s pet project will use AI to create a vibrant, virtual museum guide.<\/p>\n<p>\u201cI think we can provide more people a \u2018wow\u2019 experience \u2014 I want a copilot when I\u2019m navigating the universe,\u201d he said.<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>https:\/\/blogs.nvidia.com\/blog\/2023\/01\/24\/ai-sweden-berzelius\/<\/p>\n","protected":false},"author":0,"featured_media":2853,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[3],"tags":[],"_links":{"self":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2852"}],"collection":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/comments?post=2852"}],"version-history":[{"count":0,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2852\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media\/2853"}],"wp:attachment":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media?parent=2852"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/categories?post=2852"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/tags?post=2852"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}