lingo.lol is one of the many independent Mastodon servers you can use to participate in the fediverse.
A place for linguists, philologists, and other lovers of languages.

Server stats:

64
active users

#benchmarks

2 posts2 participants0 posts today
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p>The Best Boring <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a>: <a href="https://hachyderm.io/tags/RockyLinux10" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RockyLinux10</span></a> &amp; <a href="https://hachyderm.io/tags/AlmaLinux10" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AlmaLinux10</span></a> Performance Against <a href="https://hachyderm.io/tags/RHEL10" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RHEL10</span></a> Review<br>Testing on an AMD EPYC 9755 2P (EPYC Turin) server and using the same hardware across all tests, the performance of <a href="https://hachyderm.io/tags/RockyLinux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RockyLinux</span></a> 10 and <a href="https://hachyderm.io/tags/AlmaLinux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AlmaLinux</span></a> 10 were right on-par with <a href="https://hachyderm.io/tags/RedHat" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RedHat</span></a> <a href="https://hachyderm.io/tags/EnterpriseLinux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EnterpriseLinux</span></a> 10 itself. Hence the best kind of boring benchmarks when the performance is right on track for where it should be. <br><a href="https://www.phoronix.com/review/almalinux-10-rocky-linux-10" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/almalinux-</span><span class="invisible">10-rocky-linux-10</span></a><br><a href="https://hachyderm.io/tags/RHEL" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RHEL</span></a> <a href="https://hachyderm.io/tags/Linux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Linux</span></a></p>
UK<p><a href="https://www.europesays.com/uk/229784/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">europesays.com/uk/229784/</span><span class="invisible"></span></a> This parasite can alter your behavior, and 2 billion humans already have it <a href="https://pubeurope.com/tags/behavior" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>behavior</span></a> <a href="https://pubeurope.com/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> <a href="https://pubeurope.com/tags/Dopamine" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Dopamine</span></a> <a href="https://pubeurope.com/tags/GraphicsCard" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GraphicsCard</span></a> <a href="https://pubeurope.com/tags/Health" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Health</span></a> <a href="https://pubeurope.com/tags/humans" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>humans</span></a> <a href="https://pubeurope.com/tags/ImmuneSystem" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ImmuneSystem</span></a> <a href="https://pubeurope.com/tags/Infection" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Infection</span></a> <a href="https://pubeurope.com/tags/laptop" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>laptop</span></a> <a href="https://pubeurope.com/tags/MentalDisorders" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MentalDisorders</span></a> <a href="https://pubeurope.com/tags/netbook" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>netbook</span></a> <a href="https://pubeurope.com/tags/neuropsychiatric" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>neuropsychiatric</span></a> <a href="https://pubeurope.com/tags/notebook" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>notebook</span></a> <a href="https://pubeurope.com/tags/Parasite" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Parasite</span></a> <a href="https://pubeurope.com/tags/processor" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>processor</span></a> <a href="https://pubeurope.com/tags/reports" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>reports</span></a> <a href="https://pubeurope.com/tags/Research" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Research</span></a> <a href="https://pubeurope.com/tags/Review" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Review</span></a> <a href="https://pubeurope.com/tags/Reviews" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Reviews</span></a> <a href="https://pubeurope.com/tags/test" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>test</span></a> <a href="https://pubeurope.com/tags/tests" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>tests</span></a> <a href="https://pubeurope.com/tags/ToxoplasmaGondii" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ToxoplasmaGondii</span></a> <a href="https://pubeurope.com/tags/UK" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>UK</span></a> <a href="https://pubeurope.com/tags/UnitedKingdom" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>UnitedKingdom</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> <a href="https://hachyderm.io/tags/EPYC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPYC</span></a> <a href="https://hachyderm.io/tags/4565P" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>4565P</span></a> &amp; <a href="https://hachyderm.io/tags/4585PX" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>4585PX</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> Against <a href="https://hachyderm.io/tags/Xeon" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Xeon</span></a> <a href="https://hachyderm.io/tags/6369P" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>6369P</span></a><br>For "conventional" <a href="https://hachyderm.io/tags/server" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>server</span></a> workloads like web serving and databases, the EPYC 4005 series dominates.<br>With up to 16C/32TH, <a href="https://hachyderm.io/tags/AVX512" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AVX512</span></a>, DDR5-5600 memory and other advantages, the EPYC 4005 series is the very easy answer for those that may be looking for affordable <a href="https://hachyderm.io/tags/HPC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPC</span></a> <br>The AMD <a href="https://hachyderm.io/tags/EPYC4005" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPYC4005</span></a> series <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> deliver excellent generational uplift over the EPYC 4004 series and outright obliterating the <a href="https://hachyderm.io/tags/Xeon6300" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Xeon6300</span></a> series<br><a href="https://www.phoronix.com/review/amd-epyc-4585px-4565p-benchmarks" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/amd-epyc-4</span><span class="invisible">585px-4565p-benchmarks</span></a></p>
Thomas Wouters<p>You know how sometimes a little hobby side-project can get a bit out of hand? An unexpected performance regression on speed.python.org that only showed up on GCC 5 (and 7) led me to set up more rigorous tracking of Python performance when using different compilers. I'm still backfilling data but I think it's pretty awesome to see how much, and how consistently, free-threaded Python performance has improved since 3.13:</p><p><a href="https://github.com/Yhg1s/python-benchmarking-public" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/Yhg1s/python-benchm</span><span class="invisible">arking-public</span></a></p><p><a href="https://social.coop/tags/Python" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Python</span></a> <a href="https://social.coop/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a> <a href="https://social.coop/tags/PEP703" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>PEP703</span></a></p>
Seán Fobbe<p>🔔 New Essay 🔔 </p><p>"The Intelligent AI Coin: A Thought Experiment"</p><p>Open Access here: <a href="https://seanfobbe.com/posts/2025-02-21_intelligent-ai-coin-thought-experiment/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">seanfobbe.com/posts/2025-02-21</span><span class="invisible">_intelligent-ai-coin-thought-experiment/</span></a></p><p>Recent years have seen a concerning trend towards normalizing decisionmaking by Large Language Models (LLM), including in the adoption of legislation, the writing of judicial opinions and the routine administration of the rule of law. AI agents acting on behalf of human principals are supposed to lead us into a new age of productivity and convenience. The eloquence of AI-generated text and the narrative of super-human intelligence invite us to trust these systems more than we have trusted any human or algorithm ever before.</p><p>It is difficult to know whether a machine is actually intelligent because of problems with construct validity, plagiarism, reproducibility and transferability in AI benchmarks. Most people will either have to personally evaluate the usefulness of AI tools against the benchmark of their own lived experience or be forced to trust an expert.</p><p>To explain this conundrum I propose the Intelligent AI Coin Thought Experiment and discuss four objections: the restriction of agents to low-value decisions, making AI decisionmakers open source, adding a human-in-the-loop and the general limits of trust in human agents.</p><p><span class="h-card" translate="no"><a href="https://a.gup.pe/u/histodons" class="u-url mention" rel="nofollow noopener noreferrer" target="_blank">@<span>histodons</span></a></span> <span class="h-card" translate="no"><a href="https://a.gup.pe/u/politicalscience" class="u-url mention" rel="nofollow noopener noreferrer" target="_blank">@<span>politicalscience</span></a></span> </p><p><a href="https://fediscience.org/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://fediscience.org/tags/ArtificialIntelligence" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ArtificialIntelligence</span></a> <a href="https://fediscience.org/tags/ThoughtExperiment" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ThoughtExperiment</span></a> <a href="https://fediscience.org/tags/Philosophy" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Philosophy</span></a> <a href="https://fediscience.org/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> <a href="https://fediscience.org/tags/Validity" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Validity</span></a> <a href="https://fediscience.org/tags/Plagiarism" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Plagiarism</span></a> <a href="https://fediscience.org/tags/Reproducibility" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Reproducibility</span></a> <a href="https://fediscience.org/tags/Transferability" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Transferability</span></a> <a href="https://fediscience.org/tags/Decisionmaking" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Decisionmaking</span></a> <a href="https://fediscience.org/tags/Intelligence" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Intelligence</span></a> <a href="https://fediscience.org/tags/LLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LLM</span></a> <a href="https://fediscience.org/tags/LargeLanguageModel" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LargeLanguageModel</span></a> <a href="https://fediscience.org/tags/Algorithm" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Algorithm</span></a> <a href="https://fediscience.org/tags/Justice" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Justice</span></a> <a href="https://fediscience.org/tags/Law" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Law</span></a></p>
Linux ✅<p>📣 Guess what? Linux faster than Windows increasingly | '25 Gaming benchmarks 🐧 🎮 </p><p>Enjoy the show!</p><p>👉 <a href="https://youtu.be/D45AknAsIPw" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="">youtu.be/D45AknAsIPw</span><span class="invisible"></span></a></p><p><a href="https://linuxrocks.online/tags/Linux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Linux</span></a> <a href="https://linuxrocks.online/tags/gaming" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>gaming</span></a> <a href="https://linuxrocks.online/tags/Windows" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Windows</span></a> <a href="https://linuxrocks.online/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a> <a href="https://linuxrocks.online/tags/project2025" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>project2025</span></a> <a href="https://linuxrocks.online/tags/LinuxGaming" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LinuxGaming</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/NVIDIA" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>NVIDIA</span></a> <a href="https://hachyderm.io/tags/GeForce" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GeForce</span></a> <a href="https://hachyderm.io/tags/RTX5090" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RTX5090</span></a> <a href="https://hachyderm.io/tags/Linux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Linux</span></a> <a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a> Compute Performance <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a><br>When taking geo mean across 60+ benchmarks of <a href="https://hachyderm.io/tags/CUDA" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CUDA</span></a> / <a href="https://hachyderm.io/tags/OptiX" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>OptiX</span></a> / <a href="https://hachyderm.io/tags/OpenCL" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>OpenCL</span></a> / <a href="https://hachyderm.io/tags/Vulkan" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Vulkan</span></a> Compute, the GeForce RTX 5090 was delivering 1.42x the performance of GeForce <a href="https://hachyderm.io/tags/RTX4090" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RTX4090</span></a>. On performance-per-Watt GeForce RTX 5090 tended to deliver similar power efficiency to the RTX 4080/4090 graphics cards.<br>GeForce RTX 5090 Founders Edition was running cooler than many of the other Founders Edition graphics cards tested.<br><a href="https://www.phoronix.com/review/nvidia-geforce-rtx5090-linux" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/nvidia-gef</span><span class="invisible">orce-rtx5090-linux</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/RedHat" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RedHat</span></a> <a href="https://hachyderm.io/tags/EnterpriseLinux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EnterpriseLinux</span></a> 10 Beta Performance Looks Great - Initial <a href="https://hachyderm.io/tags/RHEL" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RHEL</span></a> 9 vs. <a href="https://hachyderm.io/tags/RHEL10" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RHEL10</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a><br><a href="https://hachyderm.io/tags/RHEL10Beta" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RHEL10Beta</span></a> is on Linux 6.11 (though <a href="https://hachyderm.io/tags/CentOSStream" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CentOSStream</span></a> 10 has moved to# Linux 6.12 LTS), GCC 14.2 is default compiler, XFS the default file-system, and there is a wealth of other package updates like moving to Python 3.12 by default and OpenJDK Java 21 as its default version. RHEL 10 also goes Wayland-only with GNOME Shell.<br>10 Beta was on average 10% faster than current RHEL 9.5<br><a href="https://www.phoronix.com/review/rhel-10-beta-benchmarks" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/rhel-10-be</span><span class="invisible">ta-benchmarks</span></a></p>
it's B! Cavello 🐝<p>Has anyone created technical <a href="https://mastodon.publicinterest.town/tags/MachineLearnjng" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MachineLearnjng</span></a> <a href="https://mastodon.publicinterest.town/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a> for generated code <a href="https://mastodon.publicinterest.town/tags/accessibility" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>accessibility</span></a>? 👀</p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> <a href="https://hachyderm.io/tags/EPYC9655" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPYC9655</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> Show The Terrific Generational Gains With 5th Gen <a href="https://hachyderm.io/tags/EPYC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPYC</span></a> Review<br>Consistent ~1.4x uplift from <a href="https://hachyderm.io/tags/EPYC9654" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EPYC9654</span></a> to EPYC 9655 was all more impressive when finding that average <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> power use only increased by about ~5% on average and only an 11% increase to the default/peak TDP. 5th Gen AMD EPYC processors continue to impress not only for raw performance over prior generation EPYC processors but also for impactful power efficiency improvements out of <a href="https://hachyderm.io/tags/Zen5" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Zen5</span></a><br><a href="https://www.phoronix.com/review/amd-epyc-9655" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/amd-epyc-9</span><span class="invisible">655</span></a></p>
Erik Jonker<p>Nice benchmark for AI models, Simple Bench.<br><a href="https://simple-bench.com/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="">simple-bench.com/</span><span class="invisible"></span></a><br><a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://mastodon.social/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a> <a href="https://mastodon.social/tags/simplebench" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>simplebench</span></a> <a href="https://mastodon.social/tags/llm" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>llm</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/Intel" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Intel</span></a> <a href="https://hachyderm.io/tags/Xeon" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Xeon</span></a> 6980P "<a href="https://hachyderm.io/tags/GraniteRapids" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GraniteRapids</span></a>" <a href="https://hachyderm.io/tags/Linux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Linux</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> Review<br>The Xeon 6980P <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> does manage to land well in front of the AMD EPYC 9684X flagship Genoa-X SKU overall along with the EPYC 9754 Bergamo processors. Generationally around 1.38x the performance with Granite Rapids compared to the prior Emerald Rapids with twice as many cores. AMD is about to launch the 5th Gen EPYC "Turin" it will be very interesting to see how those compete with <a href="https://hachyderm.io/tags/XEON6" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>XEON6</span></a>!<br><a href="https://www.phoronix.com/review/intel-xeon-6980p-performance" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/intel-xeon</span><span class="invisible">-6980p-performance</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/AWS" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AWS</span></a> <a href="https://hachyderm.io/tags/Graviton4" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Graviton4</span></a> vs. <a href="https://hachyderm.io/tags/AmpereOne" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AmpereOne</span></a> 192-Core <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> For Leading <a href="https://hachyderm.io/tags/AArch64" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AArch64</span></a> <a href="https://hachyderm.io/tags/Server" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Server</span></a> Performance<br>Particularly with <a href="https://hachyderm.io/tags/HPC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPC</span></a>-type workloads and other scientific computing scenarios the Graviton4 with its <a href="https://hachyderm.io/tags/NeoverseV2" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>NeoverseV2</span></a> cores were delivering significant leads over the custom AmpereOne cores. Its worth reiterating AmpereOne was engineered as a 2022~2023 product but it's taken now until H2'2024 for its production to really ramp up to point of seeing Oracle Cloud A2 general availability<br><a href="https://www.phoronix.com/review/ampereone-aws-graviton4" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/ampereone-</span><span class="invisible">aws-graviton4</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p>First <a href="https://hachyderm.io/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> Pitting <a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> Against <a href="https://hachyderm.io/tags/Nvidia" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Nvidia</span></a><br>Results are good in that they show <a href="https://hachyderm.io/tags/MI300X" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300X</span></a> is absolutely competitive with H100 <a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a> on one set of AI inference benchmarks, and based on our estimates of GPU and total system costs can be competitive with Nvidia’s H100 and <a href="https://hachyderm.io/tags/H200" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>H200</span></a>. But, tests only done for <a href="https://hachyderm.io/tags/Llama2" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Llama2</span></a> <a href="https://hachyderm.io/tags/LLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LLM</span></a> model from Meta with 70 billion parameters.<br>A lot will depend, on how AMD prices <a href="https://hachyderm.io/tags/MI325" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI325</span></a> later this year and how many AMD can get its partners to manufacture. <br><a href="https://www.nextplatform.com/2024/09/03/the-first-ai-benchmarks-pitting-amd-against-nvidia/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">nextplatform.com/2024/09/03/th</span><span class="invisible">e-first-ai-benchmarks-pitting-amd-against-nvidia/</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> show even an old <a href="https://hachyderm.io/tags/Nvidia" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Nvidia</span></a> <a href="https://hachyderm.io/tags/RTX3090" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>RTX3090</span></a> is enough to serve <a href="https://hachyderm.io/tags/LLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LLM</span></a> to thousands<br>For 100 concurrent users, the card delivered 12.88 tokens per second—just slightly faster than average human reading speed<br><a href="https://www.theregister.com/2024/08/23/3090_ai_benchmark/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">theregister.com/2024/08/23/309</span><span class="invisible">0_ai_benchmark/</span></a> <a href="https://hachyderm.io/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://hachyderm.io/tags/ML" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ML</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/Amazon" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Amazon</span></a>'s <a href="https://hachyderm.io/tags/Graviton" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Graviton</span></a> Has Evolved Into A Formidable <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> Contender: <a href="https://hachyderm.io/tags/Graviton1" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Graviton1</span></a> To <a href="https://hachyderm.io/tags/Graviton4" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Graviton4</span></a> <a href="https://hachyderm.io/tags/Benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Benchmarks</span></a> Review<br> Going from the Graviton3 to Graviton4 metal instance was 1.55x the performance overall, still greater than the 1.5x core difference and would be an even larger difference if isolating to only the fully multi-core benchmarks.<br>Going from the original <a href="https://hachyderm.io/tags/AWS" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AWS</span></a> Graviton1 a1.metal instance to Graviton4 metal was an incredible 10.4x!<br><a href="https://www.phoronix.com/review/aws-graviton-graviton4" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">phoronix.com/review/aws-gravit</span><span class="invisible">on-graviton4</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p>This is a longform article from Tom's Hardware, but if you want to know what AMD is giving us this year it's worth the look.</p><p><a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> deep-dives <a href="https://hachyderm.io/tags/Zen5" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Zen5</span></a> architecture — <a href="https://hachyderm.io/tags/Ryzen9000" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Ryzen9000</span></a> and <a href="https://hachyderm.io/tags/AI300" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI300</span></a> <a href="https://hachyderm.io/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a>, RDNA 3.5 <a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a>, <a href="https://hachyderm.io/tags/XDNA2" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>XDNA2</span></a>, and more <a href="https://www.tomshardware.com/pc-components/cpus/amd-deep-dives-zen-5-ryzen-9000-and-strix-point-cpu-rdna-35-gpu-and-xdna-2-architectures" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">tomshardware.com/pc-components</span><span class="invisible">/cpus/amd-deep-dives-zen-5-ryzen-9000-and-strix-point-cpu-rdna-35-gpu-and-xdna-2-architectures</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/Nvidia" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Nvidia</span></a> <a href="https://hachyderm.io/tags/GraceSuperchip" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GraceSuperchip</span></a> loses to <a href="https://hachyderm.io/tags/Intel" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Intel</span></a> <a href="https://hachyderm.io/tags/SapphireRapids" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>SapphireRapids</span></a> in <a href="https://hachyderm.io/tags/HPC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPC</span></a> performance <a href="https://hachyderm.io/tags/benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a>, promises greater efficiency<br>Barcelona <a href="https://hachyderm.io/tags/Supercomputing" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Supercomputing</span></a> and <a href="https://hachyderm.io/tags/SUNY" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>SUNY</span></a> published benchmarks showing Grace Superchip, couldn't quite match TWO 48-core Sapphire Rapids <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a>. Despite not having earth-shattering performance, Grace promises to be competitive datacenter and HPC processor thanks to its efficiency. Built on Arm architecture. A single Grace CPU comes with 72c and 480GB of LPDDR5X RAM<br><a href="https://www.tomshardware.com/pc-components/cpus/nvidia-grace-superchip-loses-to-intel-sapphire-rapids-in-hpc-performance-benchmarks-but-promises-greater-efficiency" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">tomshardware.com/pc-components</span><span class="invisible">/cpus/nvidia-grace-superchip-loses-to-intel-sapphire-rapids-in-hpc-performance-benchmarks-but-promises-greater-efficiency</span></a></p>
Hank G ☑️Intel: “Was that wrong? Should we not have done that?” <a href="https://friendica.myportal.social/search?tag=intel" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>intel</span></a> <a href="https://friendica.myportal.social/search?tag=hardware" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>hardware</span></a> <a href="https://friendica.myportal.social/search?tag=benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>benchmarks</span></a><br><a href="https://www.pcworld.com/article/2238972/intel-accused-of-inflating-cpu-benchmark-results.html" rel="nofollow noopener noreferrer" target="_blank">Intel accused of inflating CPU benchmark results</a>
gyptazySome more insights behind the <a href="https://gyptazy.ch/snac?t=starfive" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#StarFive</a> <a href="https://gyptazy.ch/snac?t=visionfive2" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#VisionFive2</a> board running on <a href="https://gyptazy.ch/snac?t=riscv64" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#RISCV64</a> architecture. Operating systems like <a href="https://gyptazy.ch/snac?t=freebsd" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#FreeBSD</a>, <a href="https://gyptazy.ch/snac?t=debian" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#Debian</a> &amp; other <a href="https://gyptazy.ch/snac?t=linux" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#Linux</a> distros, <a href="https://gyptazy.ch/snac?t=benchmarks" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#benchmarks</a> and precompiled stuff...<br><br><a href="https://gyptazy.ch/misc/collection-of-images-and-information-for-risc64-board-visionfive2/" rel="nofollow noopener noreferrer" target="_blank">https://gyptazy.ch/misc/collection-of-images-and-information-for-risc64-board-visionfive2/</a><br><br><a href="https://gyptazy.ch/snac?t=risc" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#risc</a> <a href="https://gyptazy.ch/snac?t=riscv" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#riscv</a> <a href="https://gyptazy.ch/snac?t=risc64" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#risc64</a><br>