Benjamin Carr, Ph.D. 👨🏻💻🧬<p><a href="https://hachyderm.io/tags/Intel" class="mention hashtag" rel="tag">#<span>Intel</span></a>’s <a href="https://hachyderm.io/tags/Battlemage" class="mention hashtag" rel="tag">#<span>Battlemage</span></a> Architecture<br />Battlemage is organized much like its predecessor. <a href="https://hachyderm.io/tags/Xe" class="mention hashtag" rel="tag">#<span>Xe</span></a> Cores act as a basic building block, 4 Cores grouped into a Render Slice. Xe Cores are split into XVEs, or Xe Vector engines. Intel merged pairs of Alchemist XVEs into ones that are twice as wide, completing a transition towards larger execution unit partitions. Compute is easier to utilize, cache latency improves, and weird scaling issues with global memory atomics are resolved. <br /><a href="https://chipsandcheese.com/p/intels-battlemage-architecture" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="ellipsis">chipsandcheese.com/p/intels-ba</span><span class="invisible">ttlemage-architecture</span></a></p>