Leibniz Supercomputing Centre<p>The latest <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> techniques are being discussed at <a href="https://mastodon.social/tags/ISC25" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISC25</span></a> in <a href="https://mastodon.social/tags/hamburg" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>hamburg</span></a> they are accelerating <a href="https://mastodon.social/tags/Supercomputing" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Supercomputing</span></a> and science are currently conquering this new technology and its applications. On the other hand, training AI models consumes a lot of <a href="https://mastodon.social/tags/energy" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>energy</span></a> and that needs to change: Felix Dietrich, mathematician and professor at the TUM, and his team are working on methods to calculate <a href="https://mastodon.social/tags/training" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>training</span></a> steps and thus replace them: <a href="https://www.lrz.de/en/news/detail/ai-models-replacing-training-with-math-methods" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">lrz.de/en/news/detail/ai-model</span><span class="invisible">s-replacing-training-with-math-methods</span></a></p><p><a href="https://mastodon.social/tags/connectingtheDots" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>connectingtheDots</span></a>, <a href="https://mastodon.social/tags/ISC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISC</span></a>, <a href="https://mastodon.social/tags/HPC4Germany" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HPC4Germany</span></a></p>