Habr<p>Ведущий разработчик ChatGPT и его новый проект — Безопасный Сверхинтеллект</p><p>Многие знают об Илье Суцкевере только то, что он выдающийся учёный и программист, родился в СССР, соосновал OpenAI и входит в число тех, кто в 2023 году изгнал из компании менеджера Сэма Альтмана. А когда того вернули, Суцкевер уволился по собственному желанию в новый стартап Safe Superintelligence («Безопасный Сверхинтеллект»). Илья Суцкевер действительно организовал OpenAI вместе с Маском, Брокманом, Альтманом и другими единомышленниками, причём был главным техническим гением в компании. Ведущий учёный OpenAI сыграл ключевую роль в разработке ChatGPT и других продуктов. Сейчас Илье всего 38 лет — совсем немного для звезды мировой величины.</p><p><a href="https://habr.com/ru/companies/ruvds/articles/892646/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">habr.com/ru/companies/ruvds/ar</span><span class="invisible">ticles/892646/</span></a></p><p><a href="https://zhub.link/tags/%D0%98%D0%BB%D1%8C%D1%8F_%D0%A1%D1%83%D1%86%D0%BA%D0%B5%D0%B2%D0%B5%D1%80" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Илья_Суцкевер</span></a> <a href="https://zhub.link/tags/Ilya_Sutskever" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Ilya_Sutskever</span></a> <a href="https://zhub.link/tags/OpenAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenAI</span></a> <a href="https://zhub.link/tags/10x_engineer" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>10x_engineer</span></a> <a href="https://zhub.link/tags/AlexNet" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AlexNet</span></a> <a href="https://zhub.link/tags/Safe_Superintelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Safe_Superintelligence</span></a> <a href="https://zhub.link/tags/ImageNet" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ImageNet</span></a> <a href="https://zhub.link/tags/%D0%BD%D0%B5%D0%BE%D0%BA%D0%BE%D0%B3%D0%BD%D0%B8%D1%82%D1%80%D0%BE%D0%BD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>неокогнитрон</span></a> <a href="https://zhub.link/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a> <a href="https://zhub.link/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> <a href="https://zhub.link/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a> <a href="https://zhub.link/tags/%D0%BA%D0%BE%D0%BC%D0%BF%D1%8C%D1%8E%D1%82%D0%B5%D1%80%D0%BD%D0%BE%D0%B5_%D0%B7%D1%80%D0%B5%D0%BD%D0%B8%D0%B5" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>компьютерное_зрение</span></a> <a href="https://zhub.link/tags/LeNet" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LeNet</span></a> <a href="https://zhub.link/tags/Nvidia_GTX" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Nvidia_GTX</span></a> 580 <a href="https://zhub.link/tags/DNNResearch" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DNNResearch</span></a> <a href="https://zhub.link/tags/Google_Brain" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Google_Brain</span></a> <a href="https://zhub.link/tags/%D0%90%D0%BB%D0%B5%D0%BA%D1%81_%D0%9A%D1%80%D0%B8%D0%B6%D0%B5%D0%B2%D1%81%D0%BA%D0%B8" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Алекс_Крижевски</span></a> <a href="https://zhub.link/tags/%D0%94%D0%B6%D0%B5%D1%84%D1%84%D1%80%D0%B8_%D0%A5%D0%B8%D0%BD%D1%82%D0%BE%D0%BD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Джеффри_Хинтон</span></a> <a href="https://zhub.link/tags/Seq2seq" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Seq2seq</span></a> <a href="https://zhub.link/tags/TensorFlow" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TensorFlow</span></a> <a href="https://zhub.link/tags/AlphaGo" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AlphaGo</span></a> <a href="https://zhub.link/tags/%D0%A2%D0%BE%D0%BC%D0%B0%D1%88_%D0%9C%D0%B8%D0%BA%D0%BE%D0%BB%D0%BE%D0%B2" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Томаш_Миколов</span></a> <a href="https://zhub.link/tags/Word2vec" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Word2vec</span></a> <a href="https://zhub.link/tags/fewshot_learning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>fewshot_learning</span></a> <a href="https://zhub.link/tags/%D0%BC%D0%B0%D1%88%D0%B8%D0%BD%D0%B0_%D0%91%D0%BE%D0%BB%D1%8C%D1%86%D0%BC%D0%B0%D0%BD%D0%B0" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>машина_Больцмана</span></a> <a href="https://zhub.link/tags/%D1%81%D0%B2%D0%B5%D1%80%D1%85%D0%B8%D0%BD%D1%82%D0%B5%D0%BB%D0%BB%D0%B5%D0%BA%D1%82" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>сверхинтеллект</span></a> <a href="https://zhub.link/tags/GPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPT</span></a> <a href="https://zhub.link/tags/ChatGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ChatGPT</span></a> <a href="https://zhub.link/tags/ruvds_%D1%81%D1%82%D0%B0%D1%82%D1%8C%D0%B8" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ruvds_статьи</span></a></p>
mastodon.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
The original server operated by the Mastodon gGmbH non-profit
Administered by:
Server stats:
337Kactive users
mastodon.social: About · Status · Profiles directory · Privacy policy
Mastodon: About · Get the app · Keyboard shortcuts · View source code · v4.4.0-nightly.2025-04-01
#gpgpu
1 post · 1 participant · 0 posts today
April :verified8:<p>Proposal: stop galling <a href="https://bitbang.social/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> accelerators like <a href="https://bitbang.social/tags/Nvidia" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Nvidia</span></a> Tesla (or even Quadro RTX) or <a href="https://bitbang.social/tags/AMD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AMD</span></a> Instinct "<a href="https://bitbang.social/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a>"s and start calling them "parallelizer cards", or "parallel accelerators", or the like. I will effective now.</p>
Dr. Moritz Lehmann<p>Hot Aisle's 8x AMD <a href="https://mast.hpc.social/tags/MI300X" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MI300X</span></a> server is the fastest computer I've ever tested in <a href="https://mast.hpc.social/tags/FluidX3D" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FluidX3D</span></a> <a href="https://mast.hpc.social/tags/CFD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CFD</span></a>, achieving a peak <a href="https://mast.hpc.social/tags/LBM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LBM</span></a> performance of 205 GLUPs/s, and a combined VRAM bandwidth of 23 TB/s. 🖖🤯<br>The <a href="https://mast.hpc.social/tags/RTX" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>RTX</span></a> 5090 looks like a toy in comparison.</p><p>MI300X beats even Nvidia's GH200 94GB. This marks a very fascinating inflection point in <a href="https://mast.hpc.social/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a>: <a href="https://mast.hpc.social/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a> is not the performance leader anymore. 🖖😛<br>You need a cross-vendor language like <a href="https://mast.hpc.social/tags/OpenCL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenCL</span></a> to leverage its power.</p><p>FluidX3D on <a href="https://mast.hpc.social/tags/GitHub" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GitHub</span></a>: <a href="https://github.com/ProjectPhysX/FluidX3D" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/ProjectPhysX/FluidX</span><span class="invisible">3D</span></a></p>
Giuseppe Bilotta<p>First day of the <a href="https://fediscience.org/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> course at <a href="https://fediscience.org/tags/UniCT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>UniCT</span></a>. Class is small, but students seem curious, gave me the opportunity to discuss in more details some things that usually go unmentioned. Hopefully it'll hold.</p><p>Only negative side, I had to take a longer route home because the park between my house and the university was closed 8-(</p>
Giuseppe Bilotta<p>I'm getting the material ready for my upcoming <a href="https://fediscience.org/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> course that starts on March. Even though I most probably won't get to it,I also checked my trivial <a href="https://fediscience.org/tags/SYCL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SYCL</span></a> programs. Apparently the 2025.0 version of the <a href="https://fediscience.org/tags/Intel" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Intel</span></a> <a href="https://fediscience.org/tags/OneAPI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OneAPI</span></a> <a href="https://fediscience.org/tags/DPCPP" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DPCPP</span></a> runtime doesn't like any <a href="https://fediscience.org/tags/OpenCL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenCL</span></a> platform except Intel's own (I have two other platforms that support <a href="https://fediscience.org/tags/SPIRV" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SPIRV</span></a>, so why aren't they showing up? From the documentation I can find online this should be sufficient, but apparently it's not …)</p>
🅴🆁🆄🅰 🇷🇺Это же <a href="https://www.bilibili.com/opus/1012175160065654784" rel="nofollow noopener" target="_blank">сколько заморочек</a> на NVidia и Windows'ах чтобы поиграть в Го с нейронкой? (с KataGo, аналог AlphaGo).<br><br>В таких раскладах системы на ATI/AMD и линуксах выглядят разумным выбором.<br>Раз хочется комп, чтобы играть в Го, то берёшь с такой видяхой и такой ОС, на которых меньше всего суеты с использование GPU, точнее #<a class="" href="https://hub.hubzilla.de/search?tag=GPGPU" rel="nofollow noopener" target="_blank">GPGPU</a>.<br><br>Вот по тегам <a href="https://hub.hubzilla.de/channel/erua?f=&tag=igo" rel="nofollow noopener" target="_blank">что накопилось</a> про игру в Го с компом — ничего принципиально сложного в настройке.<br>————<br>Видео, как на Windows настроить KataGo и потом использовать в #<a class="" href="https://hub.hubzilla.de/search?tag=Sabaki" rel="nofollow noopener" target="_blank">Sabaki</a> (есть portable)<br><span class="">#^</span><a class="" href="https://rutube.ru/video/de92c9e3ea5c7c1db2e1ec1b180219da/" rel="nofollow noopener" target="_blank">https://rutube.ru/video/de92c9e3ea5c7c1db2e1ec1b180219da/</a><br>И тоже самое на инглише <br><span class="">#^</span><a class="" href="https://rutube.ru/video/a94c041bc840b58115ef42134f5d845c/" rel="nofollow noopener" target="_blank">https://rutube.ru/video/a94c041bc840b58115ef42134f5d845c/</a><br>Понятно, что на youtube подобных видео ещё больше.<br><br>Видно, что используют простые варианты запуска KataGo в плане аргументов командной строки. Поскольку имеет свойство искать рядом с бинарником своим некий файл «default_model.bin.gz» содержащий ту нейросеть, что будет использоваться движком (если обратное не сказано через аргументы).<br><br>В моём случае жонглирую сетками и потому в GUI движок подключён слегка через более сложную строку:<br><code>/usr/bin/katago gtp -model /abs/path/kata1-b28c512nbt-s8032072448-d4548958859.bin.gz -human-model /abs/path/b18c384nbt-humanv0.bin.gz -config /abs/path/gtp_human_search.cfg</code><br>————<br>Может кому и смешно, а я помню времена, когда пост-советские люди приобретали себе домой шахматные компьютеры. Чтобы сами играть в шахматы и детей приучать, в домах были что-то вроде «Электроника ИМ-01». Фигуры переставлять не умел и отображал координаты хода только на табло.<br><br>Такое приобретение было непростой вещью в те времена и в тех условиях. Современность же изобилует разнообразием интеллектуальных развлечений, которые и не сильно востребованы. Однако, найдутся люди рассматривающие десктоп или ноутбук именно с точки зрения средства для игры в оффлайне, а не только в онлайн. Или для анализа своих и чужих партий опять же через нейронные сети для обучения или отработки навыков игры.<br><br>#<a class="" href="https://hub.hubzilla.de/search?tag=AMD" rel="nofollow noopener" target="_blank">AMD</a> #<a class="" href="https://hub.hubzilla.de/search?tag=ATI" rel="nofollow noopener" target="_blank">ATI</a> #<a class="" href="https://hub.hubzilla.de/search?tag=Nvidia" rel="nofollow noopener" target="_blank">Nvidia</a> #<a class="" href="https://hub.hubzilla.de/search?tag=KataGo" rel="nofollow noopener" target="_blank">KataGo</a> #<a class="" href="https://hub.hubzilla.de/search?tag=games" rel="nofollow noopener" target="_blank">games</a> #<a class="" href="https://hub.hubzilla.de/search?tag=gaming" rel="nofollow noopener" target="_blank">gaming</a> #<a class="" href="https://hub.hubzilla.de/search?tag=%D0%B3%D0%BE" rel="nofollow noopener" target="_blank">го</a> #<a class="" href="https://hub.hubzilla.de/search?tag=igo" rel="nofollow noopener" target="_blank">igo</a> #<a class="" href="https://hub.hubzilla.de/search?tag=baduk" rel="nofollow noopener" target="_blank">baduk</a> #<a class="" href="https://hub.hubzilla.de/search?tag=%D0%B1%D0%B0%D0%B4%D1%83%D0%BA" rel="nofollow noopener" target="_blank">бадук</a> #<a class="" href="https://hub.hubzilla.de/search?tag=weiqi" rel="nofollow noopener" target="_blank">weiqi</a> #<a class="" href="https://hub.hubzilla.de/search?tag=%D0%B2%D1%8D%D0%B9%D1%86%D0%B8" rel="nofollow noopener" target="_blank">вэйци</a> #<a class="" href="https://hub.hubzilla.de/search?tag=lang_ru" rel="nofollow noopener" target="_blank">lang_ru</a> @<a href="https://3zi.ru/@Russia" rel="nofollow noopener" target="_blank">Russia</a>
gigapixel<p>So I found <a href="https://github.com/tracel-ai/cubecl" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">github.com/tracel-ai/cubecl</span><span class="invisible"></span></a> which allows <a href="https://mathstodon.xyz/tags/gpgpu" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>gpgpu</span></a> in <a href="https://mathstodon.xyz/tags/rust" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>rust</span></a>. Already using it to calculate some determinants for triangulations. Wondering if it can be leveraged to build a numeric <a href="https://mathstodon.xyz/tags/pde" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>pde</span></a> solver</p>
Anthony Cowley<p>This look at how many HLSL instructions different variations of an endian swap compile to with AMD tooling is... well, frankly, it's upsetting. All variations are similar, but compile down to anywhere from 1 to 13 IR ops. Contorting code to trigger desired compilation paths is familar to many GHC Haskellers, but it's an incredible deterrent to prioritizing performance. <a href="https://martinfullerblog.wordpress.com/2025/01/13/massaging-the-shader-compiler-to-emit-optimum-instructions/" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="ellipsis">martinfullerblog.wordpress.com</span><span class="invisible">/2025/01/13/massaging-the-shader-compiler-to-emit-optimum-instructions/</span></a></p><p><a href="https://mastodon.social/tags/gpgpu" class="mention hashtag" rel="tag">#<span>gpgpu</span></a></p>
Dr. Moritz Lehmann<p><span class="h-card" translate="no"><a href="https://hachyderm.io/@BenjaminHCCarr" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>BenjaminHCCarr</span></a></span> another article on <a href="https://mast.hpc.social/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a> code portability where people put their heads in the sand and pretend very hard that <a href="https://mast.hpc.social/tags/OpenCL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenCL</span></a> doesn't exist...<br>OpenCL has solved <a href="https://mast.hpc.social/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> cross-compatibility 16 years ago already and today is in better shape than ever.</p>
Kevin Karhan :verified:<p><span class="h-card" translate="no"><a href="https://mk.absturztau.be/@enigmatico" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>enigmatico</span></a></span> <span class="h-card" translate="no"><a href="https://udongein.xyz/users/lispi314" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>lispi314</span></a></span> <span class="h-card" translate="no"><a href="https://ublog.kimapr.net/users/kimapr" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>kimapr</span></a></span> <span class="h-card" translate="no"><a href="https://transfem.social/@bunnybeam" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>bunnybeam</span></a></span> case in point:</p><ul><li><p><a href="https://infosec.space/tags/Bloatedness" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Bloatedness</span></a> was <a href="https://mk.absturztau.be/notes/a2x3lafb4lky00q6" rel="nofollow noopener" target="_blank">the original post</a> topic and yes, due to <a href="https://infosec.space/tags/TechBros" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TechBros</span></a> <em>"<a href="https://infosec.space/tags/BuildFastBreakThings" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BuildFastBreakThings</span></a>"</em> mentality, <a href="https://infosec.space/tags/Bloatware" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Bloatware</span></a> is increasing given that a shitty bloated 50+MB <em>"<a href="https://infosec.space/tags/WebApp" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WebApp</span></a>"</em> with like <a href="https://nwjs.io" rel="nofollow noopener" target="_blank">nw.js</a> is easy to slap together (and yes I did so myself!) than to put in way more thought and effort (as you can see on the slow progression of <a href="https://infosec.space/@OS1337" rel="nofollow noopener" target="_blank">OS/1337</a>...</p></li><li><p>Yes, <a href="https://infosec.space/tags/Accessibility" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Accessibility</span></a> is something that needs to be taken more seriously and it's good to see that there's at least some attemots at making <a href="https://infosec.space/tags/accessibility" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>accessibility</span></a> mandatory (at least in <a href="https://infosec.space/tags/Germany" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Germany</span></a>, where I know from some insider that a big telco is investing a lot in that!) for a growng number of industries and websites...</p></li><li><p>And whilst one can slap an <a href="https://infosec.space/tags/RTX5090" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>RTX5090</span></a> on any laptop that has a fully-functional <a href="https://infosec.space/tags/ExpressCard" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ExpressCard</span></a> slot (with <a href="https://infosec.space/tags/PCIe" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PCIe</span></a> interface, using some janky adaptors!) that'll certainly not make sense beyond some <a href="https://infosec.space/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a> or other <a href="https://infosec.space/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a>-style workloads as it's bottlenecked to a single PCIe lane of 2.0 (500MB/s) or just 1.0a(250MB/s) speeds.</p></li></ul><p>Needless to say there is a need to <em>THINN DOWN</em> things cuz the current speed of <a href="https://infosec.space/tags/Enshittifcation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Enshittifcation</span></a> and bloatedness combined with <a href="https://infosec.space/tags/AntiRepairDesign" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AntiRepairDesign</span></a> and <em>overpriced yet worse <a href="https://infosec.space/tags/tech" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>tech</span></a></em> in general makes it unsustainable for an ever increasing population!</p><ul><li><p>Not everyone wants (or even can!) indebt themselves just to have a phone or laptop!</p></li><li></li><li></li></ul><p>Should we aim for more <em>"<a href="https://infosec.space/tags/FrugslComputing" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FrugslComputing</span></a>"</em>? </p><ul><li>Abdolutely!</li></ul><p>Is it realistic to expect things to be in a perfectly accessible TUI that ebery screenreader can handle?</p><ul><li>No!</li></ul><p>That being said the apathy of consumers is real, and very frustrating:</p><p>People get nudged into accepting all the bs and it really pisses me off because they want me to look like ab outsider / asshole for not submitting to <a href="https://infosec.space/tags/consumerism" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>consumerism</span></a> and <a href="https://infosec.space/tags/unsustainable" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>unsustainable</span></a> shite...</p>
Mimetik<p>Still work in progress: debugging a reaction-diffusion compute shader for a GPU generated mesh.</p><p><a href="https://genart.social/tags/openrndr" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>openrndr</span></a> <a href="https://genart.social/tags/cellforms" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>cellforms</span></a> <a href="https://genart.social/tags/reactiondiffusion" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>reactiondiffusion</span></a> <a href="https://genart.social/tags/gpgpu" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>gpgpu</span></a></p>
Oblomov<p>Even better, in the afternoon I managed to find a workaround for my <a href="https://sociale.network/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> software building but hanging when trying to run it, which seems to be related to an issue with some versions of the <a href="https://sociale.network/tags/AMD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AMD</span></a> software stack and many integrated GPUs, not just the <a href="https://sociale.network/tags/SteamDeck" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SteamDeck</span></a> specifically. So exporting the HSA_ENABLE_SDMA=0 environment vriable was sufficient to get my software running again. I'm dropping the information here in case others find it useful.</p><p><a href="https://sociale.network/tags/ROCm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ROCm</span></a> <a href="https://sociale.network/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a> <a href="https://sociale.network/tags/APU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>APU</span></a> <a href="https://sociale.network/tags/HIP" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HIP</span></a></p><p>2/2</p>
Tech Cybersecurity Nieuws<p>Intel's nieuwe ai accelerator: jaguar shores ontmaskerd <a href="https://www.trendingtech.news/trending-news/2024/11/49225/intel-s-nieuwe-ai-accelerator-jaguar-shores-ontmaskerd" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">trendingtech.news/trending-new</span><span class="invisible">s/2024/11/49225/intel-s-nieuwe-ai-accelerator-jaguar-shores-ontmaskerd</span></a> <a href="https://mastodon.nl/tags/Intel" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Intel</span></a> <a href="https://mastodon.nl/tags/Jaguar" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Jaguar</span></a> Shores <a href="https://mastodon.nl/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> Accelerator <a href="https://mastodon.nl/tags/Falcon" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Falcon</span></a> Shores <a href="https://mastodon.nl/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> <a href="https://mastodon.nl/tags/Trending" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Trending</span></a> <a href="https://mastodon.nl/tags/News" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>News</span></a> <a href="https://mastodon.nl/tags/Nieuws" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Nieuws</span></a></p>
Chapel Programming Language<p>GPU programming can be challenging. Join Engin Kayraklioglu’s live demo today (October 31st) at 10AM Pacific to see how the Chapel programming language makes it easier to program GPUs.</p><p>Find out more on the Upcoming Events page: <a href="https://chapel-lang.org/events.html" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="">chapel-lang.org/events.html</span><span class="invisible"></span></a></p><p><a href="https://mastodon.social/tags/GPU" class="mention hashtag" rel="tag">#<span>GPU</span></a> <a href="https://mastodon.social/tags/GPGPU" class="mention hashtag" rel="tag">#<span>GPGPU</span></a> <a href="https://mastodon.social/tags/ChapelLang" class="mention hashtag" rel="tag">#<span>ChapelLang</span></a> <a href="https://mastodon.social/tags/CUDA" class="mention hashtag" rel="tag">#<span>CUDA</span></a></p>
Giuseppe Bilotta<p>It's out, if anyone is curious</p><p> <a href="https://doi.org/10.1002/cpe.8313" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">doi.org/10.1002/cpe.8313</span><span class="invisible"></span></a></p><p>This is a “how to” guide. <a href="https://fediscience.org/tags/GPUSPH" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPUSPH</span></a>, as the name suggests, was designed from the ground up to run on <a href="https://fediscience.org/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a> (w/ <a href="https://fediscience.org/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a>, for historical reasons). We wrote a CPU version a long time ago for a publication that required a comparison, but it was never maintained. In 2021, I finally took the plunge, and taking inspiration from <a href="https://fediscience.org/tags/SYCL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SYCL</span></a>, adapted the device code in functor form, so that it could be “trivially” compiled for CPU as well.</p><p><a href="https://fediscience.org/tags/HPC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HPC</span></a> <a href="https://fediscience.org/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a></p>
Giuseppe Bilotta<p>I love the smell of burning plastic when running a heavy-duty <a href="https://fediscience.org/tags/CFD" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CFD</span></a> simulation on my laptop's <a href="https://fediscience.org/tags/GPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPU</span></a> </p><p><a href="https://fediscience.org/tags/humor" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>humor</span></a> <a href="https://fediscience.org/tags/humour" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>humour</span></a> <a href="https://fediscience.org/tags/HPC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HPC</span></a> <a href="https://fediscience.org/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a></p>
:rss: Qiita - 人気の記事<p>Intel内蔵GPUをoneAPI Level Zeroで直接叩いてみる（環境構築編）<br><a href="https://qiita.com/ysktrad/items/437a464bbae7ea42dc7e?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">qiita.com/ysktrad/items/437a46</span><span class="invisible">4bbae7ea42dc7e?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items</span></a></p><p><a href="https://rss-mstdn.studiofreesia.com/tags/qiita" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>qiita</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/C" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>C</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/Ubuntu" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Ubuntu</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/setup" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>setup</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/IntelGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>IntelGPU</span></a></p>
Anthony Cowley<p>Here’s hoping that the transition of <a href="https://mastodon.social/tags/Rust" class="mention hashtag" rel="tag">#<span>Rust</span></a> <a href="https://mastodon.social/tags/GPU" class="mention hashtag" rel="tag">#<span>GPU</span></a> to community ownership goes well! The intention to focus on <a href="https://mastodon.social/tags/GPGPU" class="mention hashtag" rel="tag">#<span>GPGPU</span></a> is more than welcome, as I feel the development of some GPU programming ecosystems has been held back by a too-narrow focus on traditional GPU graphics techniques. The dream is to be able to write CUDA-like Rust that can target hardware from multiple vendors!</p><p><a href="https://rust-gpu.github.io/blog/transition-announcement/" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="ellipsis">rust-gpu.github.io/blog/transi</span><span class="invisible">tion-announcement/</span></a></p>
GeekProjects News<p>CUDA, But Make It AMD <a href="https://hackaday.com/2024/07/16/cuda-but-make-it-amd/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">hackaday.com/2024/07/16/cuda-b</span><span class="invisible">ut-make-it-amd/</span></a> <a href="https://geekprojects.com/tags/generalpurposegpu" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>generalpurposegpu</span></a> <a href="https://geekprojects.com/tags/MachineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MachineLearning</span></a> <a href="https://geekprojects.com/tags/MiscHacks" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MiscHacks</span></a> <a href="https://geekprojects.com/tags/radeon" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>radeon</span></a> <a href="https://geekprojects.com/tags/gpgpu" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>gpgpu</span></a> <a href="https://geekprojects.com/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a> <a href="https://geekprojects.com/tags/amd" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>amd</span></a> <a href="https://geekprojects.com/tags/ATI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ATI</span></a></p>
Karsten Schmidt<p>Uploaded a new demo/example showing how to perform GPU-side data reductions using <a href="https://thi.ng/shader-ast" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">thi.ng/shader-ast</span><span class="invisible"></span></a> & <a href="https://thi.ng/webgl" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">thi.ng/webgl</span><span class="invisible"></span></a> multi-pass pipeline. Arbitrary reduction functions supported. If there's interest, this could be expanded & packaged up as library... 90% of this example is boiler plate, 9.9% benchmarking & debug outputs...</p><p>Demo:<br><a href="https://demo.thi.ng/umbrella/gpgpu-reduce/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">demo.thi.ng/umbrella/gpgpu-red</span><span class="invisible">uce/</span></a></p><p>Source code:<br><a href="https://github.com/thi-ng/umbrella/blob/develop/examples/gpgpu-reduce/src/index.ts" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/thi-ng/umbrella/blo</span><span class="invisible">b/develop/examples/gpgpu-reduce/src/index.ts</span></a></p><p>Readme w/ benchmark results:<br><a href="https://github.com/thi-ng/umbrella/tree/develop/examples/gpgpu-reduce" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/thi-ng/umbrella/tre</span><span class="invisible">e/develop/examples/gpgpu-reduce</span></a></p><p>Related discussion:<br><a href="https://github.com/thi-ng/umbrella/issues/478" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/thi-ng/umbrella/iss</span><span class="invisible">ues/478</span></a></p><p><a href="https://mastodon.thi.ng/tags/ThingUmbrella" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ThingUmbrella</span></a> <a href="https://mastodon.thi.ng/tags/WebGL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WebGL</span></a> <a href="https://mastodon.thi.ng/tags/ShaderAST" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ShaderAST</span></a> <a href="https://mastodon.thi.ng/tags/GPGPU" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPGPU</span></a> <a href="https://mastodon.thi.ng/tags/TypeScript" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TypeScript</span></a> <a href="https://mastodon.thi.ng/tags/JavaScript" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>JavaScript</span></a></p>
ExploreLive feeds
Mastodon is the best way to keep up with what's happening.
Follow anyone across the fediverse and see it all in chronological order. No algorithms, ads, or clickbait in sight.
Create accountLoginDrag & drop to upload