{"id":96685,"date":"2025-08-20T11:30:09","date_gmt":"2025-08-20T11:30:09","guid":{"rendered":"https:\/\/www.newsbeep.com\/us\/96685\/"},"modified":"2025-08-20T11:30:09","modified_gmt":"2025-08-20T11:30:09","slug":"terakraft-and-neurophos-to-develop-sustainable-ai-computing","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/us\/96685\/","title":{"rendered":"Terakraft and Neurophos to develop sustainable AI computing"},"content":{"rendered":"<p>With artificial intelligence having a massive transformational impact, the demand for high performance computing infrastructure is growing rapidly. One of the biggest challenges facing the computing infrastructure industry, however, is the immense energy consumption required to power and cool these systems.<\/p>\n<p>Terakraft and Neurophos are coming together to demonstrate how greener data centres and breakthrough hardware can work together to deliver a more sustainable path going forward.<\/p>\n<p>Located in Norway, Terakraft uses power entirely sourced by renewable hydropower and cooled with natural lake water and has achieved a Power Usage Effectiveness (PUE) below 1.1 \u2013 among the best metrics in the industry. By repurposing an existing hydropower plant built with reinforced concrete, Terakraft avoids additional embodied carbon and has been able to set a new benchmark for low-emission operations.<\/p>\n<p>Neurophos\u2019s AI hardware has been developed using proprietary optical processing units which are 100x more energy efficient than many current GPUs for the same workload.<\/p>\n<p>According to Neurophis, by harnessing the efficiency of photonics and scaling properties from metamaterials, it has fabricated, demonstrated, and characterized its miniaturisation of optical modulators by a factor of 10,000x, enabling future optical processing units to deliver the compute power of 100 GPUs while consuming the equivalent of 1% of the energy \u2013 claims that have been validated by end-to-end simulation results.<\/p>\n<p>Its compute-in-memory architecture, which has been inspired by the human brain\u2019s efficiency, allows for accelerated processing speeds and density, making ultra-efficient, large-scale AI inference practical and scalable.<\/p>\n<p>The companies plan to host a pilot as part of a commercial early access program in 2027 for Neurophos\u2019 accelerated AI inference platform. The project will provide a real-world proving ground for sustainable, ultra-efficient compute, complementing today\u2019s GPU-driven systems with experimental next-generation photonic hardware.<\/p>\n<p>\u201cBy hosting Neurophos\u2019 ultra-efficient optical chips in our green data centre for select enterprise clients, we\u2019re not only reducing our carbon footprint but raising the bar for energy-efficient AI infrastructure. Our mission has always been to power the future responsibly, and this collaboration brings that vision to life,\u201d said Giorgio Sbriglia, Chairman of the Board of Terakraft.<\/p>\n<p>\u201cTerakraft\u2019s commitment to renewable energy and innovative technologies aligns perfectly with our mission to democratise high-performance AI. By deploying our 100x more efficient inference chips in Terakraft\u2019s green data centre, we\u2019re proving that AI\u2019s exponential growth can be achieved sustainably, together,\u201d said Patrick Bowen, CEO and Founder of Neurophos.<\/p>\n<p>This collaboration is described as an innovative step forward in accelerated computing infrastructure and will demonstrate a viable path for AI growth that is both economically and environmentally sustainable, paving the way for global adoption of green AI data centres and hardware.<\/p>\n","protected":false},"excerpt":{"rendered":"With artificial intelligence having a massive transformational impact, the demand for high performance computing infrastructure is growing rapidly.&hellip;\n","protected":false},"author":2,"featured_media":96686,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[46],"tags":[65302,191,10377,65303,62542,5568,6215,74],"class_list":{"0":"post-96685","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-computing","8":"tag-artificial-intellgence","9":"tag-computing","10":"tag-data-centres","11":"tag-green-power","12":"tag-photonics","13":"tag-power","14":"tag-semiconductors","15":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/96685","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/comments?post=96685"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/96685\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media\/96686"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media?parent=96685"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/categories?post=96685"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/tags?post=96685"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}