Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
en.osm.town is one of the many independent Mastodon servers you can use to participate in the fediverse.
An independent, community of OpenStreetMap people on the Fediverse/Mastodon. Funding graciously provided by the OpenStreetMap Foundation.

Administered by:

Server stats:

256
active users

en.osm.town: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#mi300a

0 posts0 participants0 posts today
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p>Sizing up <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a>’s <a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a><br>It’s well ahead of <a href="https://hachyderm.io/tags/Nvidia" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Nvidia</span></a>’s <a href="https://hachyderm.io/tags/H100" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>H100</span></a> PCIe for just about every major category of 32- or 64-bit operations. MI300A can achieve 113.2 TFLOPS of <a href="https://hachyderm.io/tags/FP32" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>FP32</span></a> throughput, with each FMA counting as two floating point operations. For comparison, H100 PCIe achieved 49.3 TFLOPS in same test.<br><a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> cut down <a href="https://hachyderm.io/tags/MI300X" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300X</span></a>’s GPU to create MI300A. 24 <a href="https://hachyderm.io/tags/Zen4" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Zen4</span></a> cores is a lot of <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> power, and occupies one quadrant on the MI300 chip. But MI300’s main attraction is still the GPU.<br><a href="https://chipsandcheese.com/p/sizing-up-mi300as-gpu" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">chipsandcheese.com/p/sizing-up</span><span class="invisible">-mi300as-gpu</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/Germany" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Germany</span></a> unleashes <a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a>-powered Hunter <a href="https://hachyderm.io/tags/supercomputer" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>supercomputer</span></a><br>€15 million system to serve as testbed for larger Herder supercomputer coming in 2027<br>Built by <a href="https://hachyderm.io/tags/HewlettPackardEnterprise" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HewlettPackardEnterprise</span></a> (<a href="https://hachyderm.io/tags/HPE" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPE</span></a>), Hunter is based on a <a href="https://hachyderm.io/tags/Cray" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Cray</span></a> EX4000 platform and powered by a combination of AMD Instinct <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> accelerated processing units (<a href="https://hachyderm.io/tags/APU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>APU</span></a>) and <a href="https://hachyderm.io/tags/Epyc" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Epyc</span></a> Genoa <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a>. <br><a href="https://www.theregister.com/2025/01/17/hlrs_supercomputer_hunter/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">theregister.com/2025/01/17/hlr</span><span class="invisible">s_supercomputer_hunter/</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/HPE" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPE</span></a> <a href="https://hachyderm.io/tags/Cray" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Cray</span></a> <a href="https://hachyderm.io/tags/EX4000" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EX4000</span></a> 'Hunter' <a href="https://hachyderm.io/tags/Supercomputer" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Supercomputer</span></a> Now in Operation at <a href="https://hachyderm.io/tags/HLRS" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HLRS</span></a> Powered by <a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> APUs<br>With a theoretical peak performance of 48.1 petaflops (48.1 quadrillion floating point operations), Hunter’s speed is nearly double that of HLRS’s previous flagship supercomputer, called Hawk. Hunter is based on the AMD Instinct <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> accelerated processing unit (<a href="https://hachyderm.io/tags/APU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>APU</span></a>), which combines CPUs, GPU accelerators, and high bandwidth memory in a single package. <br><a href="https://insidehpc.com/2025/01/hpe-hunter-supercomputer-now-in-operation-at-hlrs-powered-by-amd-apus/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">insidehpc.com/2025/01/hpe-hunt</span><span class="invisible">er-supercomputer-now-in-operation-at-hlrs-powered-by-amd-apus/</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p><a href="https://hachyderm.io/tags/ElCapitan" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ElCapitan</span></a> Towers Above the <a href="https://hachyderm.io/tags/Top500" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Top500</span></a> in a Big <a href="https://hachyderm.io/tags/HPE" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPE</span></a> Win using <a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a> Instinct <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> system. With sustained <a href="https://hachyderm.io/tags/HPL" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPL</span></a> of 1.742EF and peak speed of 2.79EF of <a href="https://hachyderm.io/tags/FP64" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>FP64</span></a>, this is a big jump over previous generation systems. El Capitan has a unique architecture as it uses <a href="https://hachyderm.io/tags/APU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>APU</span></a> combining <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> plus <a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a> onto a package with high-bandwidth memory. Over 44,000 of these MI300A APUs are then packed into the HPE <a href="https://hachyderm.io/tags/Cray" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Cray</span></a> Shasta liquid cooled platform, and connected via the Slingshot interconnect. <br><a href="https://www.servethehome.com/el-capitan-towers-above-the-top500-in-a-big-hpe-and-amd-win/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">servethehome.com/el-capitan-to</span><span class="invisible">wers-above-the-top500-in-a-big-hpe-and-amd-win/</span></a></p>
Benjamin Carr, Ph.D. 👨🏻‍💻🧬<p>All the tech and history that culminates in the MI300A accelerator!</p><p><a href="https://hachyderm.io/tags/AMD" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AMD</span></a>’s Long And Winding Road To The Hybrid <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a>-<a href="https://hachyderm.io/tags/GPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>GPU</span></a> Instinct <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a><br><a href="https://www.nextplatform.com/2024/07/17/amds-long-and-winding-road-to-the-hybrid-cpu-gpu-instinct-mi300a/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">nextplatform.com/2024/07/17/am</span><span class="invisible">ds-long-and-winding-road-to-the-hybrid-cpu-gpu-instinct-mi300a/</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/Instinct" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Instinct</span></a> <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> <a href="https://hachyderm.io/tags/APU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>APU</span></a> With <a href="https://hachyderm.io/tags/CDNA3" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CDNA3</span></a> <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/Zen4" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Zen4</span></a> <a href="https://hachyderm.io/tags/CPU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CPU</span></a> &amp; <a href="https://hachyderm.io/tags/UnifiedMemory" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>UnifiedMemory</span></a> Offers Up To 4x Speedup Versus Discrete GPUs In <a href="https://hachyderm.io/tags/HPC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPC</span></a><br>Since the AMD Instinct MI300A accelerator uses a unified <a href="https://hachyderm.io/tags/HBM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HBM</span></a> interface, it eliminates the need for data replication and does not require a <a href="https://hachyderm.io/tags/programming" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>programming</span></a> distinction between the host and the device memory spaces. <br><a href="https://wccftech.com/amd-instinct-mi300a-apu-cdna-3-gpu-zen-4-cpu-unified-memory-up-to-4x-speedup-versus-discrete-gpus/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">wccftech.com/amd-instinct-mi30</span><span class="invisible">0a-apu-cdna-3-gpu-zen-4-cpu-unified-memory-up-to-4x-speedup-versus-discrete-gpus/</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> bets demand for its <a href="https://hachyderm.io/tags/MI300" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300</span></a> accelerator will balance dips across other product lines<br>CEO <a href="https://hachyderm.io/tags/LisaSu" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LisaSu</span></a> boasted AMD now predicts <a href="https://hachyderm.io/tags/datacenter" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>datacenter</span></a> GPUs – this includes the <a href="https://hachyderm.io/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> – to see consistent sequential growth over the next several quarters, driving revenues of $3.5 billion in 2024 alone. Tthe MI300 represents AMD’s first compelling contender in the <a href="https://hachyderm.io/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> arena, with its previous-gen <a href="https://hachyderm.io/tags/MI200" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI200</span></a> parts primarily designed for <a href="https://hachyderm.io/tags/highperformancecomputing" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>highperformancecomputing</span></a> (<a href="https://hachyderm.io/tags/hpc" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>hpc</span></a>) deployments <br><a href="https://www.theregister.com/2024/01/31/amd_q4_2023/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">theregister.com/2024/01/31/amd</span><span class="invisible">_q4_2023/</span></a> <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/APU" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>APU</span></a></p>
HPC Guru<p>The University of Stuttgart and <br>HPE to build two new <a href="https://mastodon.social/tags/supercomputers" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>supercomputers</span></a> at HLRS</p><p>o Hunter replaces Hawk - HPE/Cray based on AMD <a href="https://mastodon.social/tags/MI300A" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>MI300A</span></a> APU</p><p>o Herder will be an <a href="https://mastodon.social/tags/Exascale" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Exascale</span></a> system for 2027 - final config by YE 2025</p><p><a href="https://hlrs.de/news/detail/exascale-supercomputing-is-coming-to-stuttgart" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">hlrs.de/news/detail/exascale-s</span><span class="invisible">upercomputing-is-coming-to-stuttgart</span></a></p><p><a href="https://mastodon.social/tags/HPC" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HPC</span></a> <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> via <br><span class="h-card" translate="no"><a href="https://xn--baw-joa.social/@Uni_Stuttgart" class="u-url mention" rel="nofollow noopener noreferrer" target="_blank">@<span>Uni_Stuttgart</span></a></span></p>
ExploreLive feeds
About