Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
mathstodon.xyz is one of the many independent Mastodon servers you can use to participate in the fediverse.
A Mastodon instance for maths people. We have LaTeX rendering in the web interface!

Administered by:

Server stats:

3K
active users

mathstodon.xyz: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#dftk

0 posts0 participants0 posts today
Michael Herbst<p>Just delivered a talk on algorithmic differentiation in our <a href="https://social.epfl.ch/tags/dftk" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>dftk</span></a> density-functional theory code at the <a href="https://social.epfl.ch/tags/cecam" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>cecam</span></a> workshop on <a href="https://social.epfl.ch/tags/dft" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>dft</span></a> and <a href="https://social.epfl.ch/tags/ai" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ai</span></a> </p><p>Slides: <a href="https://michael-herbst.com/talks/2025.03.04_DFT_AI_Learning_each_other.pdf" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">michael-herbst.com/talks/2025.</span><span class="invisible">03.04_DFT_AI_Learning_each_other.pdf</span></a><br><a href="https://social.epfl.ch/tags/julialang" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>julialang</span></a> demo: <a href="https://michael-herbst.com/talks/2025.03.04_DFT_AI_Learning_each_other_design.html" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">michael-herbst.com/talks/2025.</span><span class="invisible">03.04_DFT_AI_Learning_each_other_design.html</span></a></p>
Michael Herbst<p>We just released <a href="https://social.epfl.ch/tags/DFTK" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>DFTK</span></a> version 0.7.10: <a href="http://dftk.org/releases" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">http://</span><span class="">dftk.org/releases</span><span class="invisible"></span></a> with an important bug fix for meta-GGA band structure computations and some details how to use DFTK on GPUs. Either <a href="https://social.epfl.ch/tags/cuda" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>cuda</span></a> (<a href="https://social.epfl.ch/tags/nvidia" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>nvidia</span></a> GPUs) or <a href="https://social.epfl.ch/tags/rocm" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>rocm</span></a> (<a href="https://social.epfl.ch/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> GPUs) are supported.</p>
MatMat<p><span class="h-card" translate="no"><a href="https://social.epfl.ch/@schmitz" class="u-url mention" rel="nofollow noopener noreferrer" target="_blank">@<span>schmitz</span></a></span> (left) explaining his recent work on making <a href="https://social.epfl.ch/tags/dftk" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>dftk</span></a> algorithmically <a href="https://social.epfl.ch/tags/differentiable" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>differentiable</span></a> at the <a href="https://social.epfl.ch/tags/cecam" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>cecam</span></a> workshop on <a href="https://social.epfl.ch/tags/dft" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>dft</span></a> and <a href="https://social.epfl.ch/tags/ai" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ai</span></a> (<a href="https://www.cecam.org/workshop-details/1281" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">cecam.org/workshop-details/128</span><span class="invisible">1</span></a>). With his work derivatives of key density-functional theory quantities like forces or band structures wrt. model parameters can now be easily computed.</p>
Michael Herbst<p>Hello <a href="https://social.epfl.ch/tags/fediverse" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>fediverse</span></a> <a href="https://social.epfl.ch/tags/introduction" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>introduction</span></a></p><p>I'm Michael, professor in the institutes of <a href="https://social.epfl.ch/tags/mathematics" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>mathematics</span></a> and <a href="https://social.epfl.ch/tags/materials" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>materials</span></a> science and head of the <span class="h-card" translate="no"><a href="https://social.epfl.ch/@MatMat" class="u-url mention" rel="nofollow noopener noreferrer" target="_blank">@<span>MatMat</span></a></span> group at <a href="https://social.epfl.ch/tags/EPFL" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPFL</span></a>.</p><p>I work on the <a href="https://social.epfl.ch/tags/atomistic" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>atomistic</span></a> simulations of materials, mainly density-functional theory (DFT) methods, understanding <a href="https://social.epfl.ch/tags/simulation" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>simulation</span></a> errors and <a href="https://social.epfl.ch/tags/uncertainties" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>uncertainties</span></a> in predicted materials properties.</p><p>I use techniques from <br><a href="https://social.epfl.ch/tags/physics" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>physics</span></a> <a href="https://social.epfl.ch/tags/computerscience" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>computerscience</span></a> <a href="https://social.epfl.ch/tags/machinelearning" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>machinelearning</span></a> and <br>develop related <a href="https://social.epfl.ch/tags/julialang" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>julialang</span></a> packages such as the density-functional toolkit (<a href="https://social.epfl.ch/tags/dftk" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>dftk</span></a>).</p>
SearchLive feeds
About