Deploying to gh-pages from @ 2f96129 🚀

index.html

@@ -195,6 +195,8 @@ <h2 id="community"><a href="#community" class="header-anchor">Community</a></h2>
 </li>
 <li><p>Julia Slack &#40;<a href="https://slackinvite.julialang.org/">register here</a>&#41;, on the <a href="https://julialang.slack.com/messages/C689Y34LE/">#gpu channel</a></p>
 </li>
+<li><p><a href="https://mit.zoom.us/j/92410737378?pwd&#61;MWkrSkIzMEtRWkZHdm1XQmpNbm90UT09">JuliaGPU office hours</a>, every other week at 2PM CET &#40;check the <a href="https://julialang.org/community/#events">Julia community calendar</a> for more details&#41;.</p>
+</li>
 </ul>
 <!-- CONTENT ENDS HERE -->
       </main>

other/index.html

@@ -138,7 +138,7 @@
 <h1 id="other"><a href="#other" class="header-anchor">Other</a></h1>
 <p>Several other back-ends exist, not all of them with the same level of polish or support as the NVIDIA and AMD back-ends.</p>
 <h2 id="opencl"><a href="#opencl" class="header-anchor">OpenCL</a></h2>
-<p>Programming OpenCL GPUs in Julia is much more limited than other supported platforms. On recent versions of Julia, only <a href="https://github.com/JuliaGPU/OpenCL.jl">OpenCL.jl</a> is available. This package can be used to compile and execute GPU kernels written in OpenCL C.</p>
+<p>Support for OpenCL is available through the <a href="https://github.com/JuliaGPU/OpenCL.jl">OpenCL.jl</a> package, which is currently undergoing a rewrite to enable native support for Julia.</p>
 <h2 id="arrayfire"><a href="#arrayfire" class="header-anchor">ArrayFire</a></h2>
 <p>ArrayFire is a general-purpose software library that targets CPUs, GPUs, and other accelerator hardware. The <a href="https://github.com/JuliaGPU/ArrayFire.jl">ArrayFire.jl</a> package provides a Julia interface to this library, and makes it possible to program accelerators using an array abstraction built on the ArrayFire library.</p>
 <h2 id="sx-aurora"><a href="#sx-aurora" class="header-anchor">SX-Aurora</a></h2>

previews/PR46/post/2024-09-18-cuda_5.5/index.html → post/2024-09-18-cuda_5.5/index.html

@@ -4,10 +4,10 @@
   <meta charset="UTF-8">
   <meta name="viewport" content="width=device-width,initial-scale=1,shrink-to-fit=no">
   <meta content="index, follow" name=robots>
-  <link rel="icon" href="/previews/PR46/assets/favicon.ico">
-  <link rel="alternate" type="application/rss+xml" href="/previews/PR46/post/index.xml" title="RSS Feed for JuliaGPU">
+  <link rel="icon" href="/assets/favicon.ico">
+  <link rel="alternate" type="application/rss+xml" href="/post/index.xml" title="RSS Feed for JuliaGPU">
 
-  <link rel="stylesheet" href="/previews/PR46/css/bootstrap.min.css">
+  <link rel="stylesheet" href="/css/bootstrap.min.css">
 
 
 
@@ -108,21 +108,21 @@
   <div id=nav-border class=container>
     <nav class="navbar navbar-expand-lg navbar-light justify-content-center">
         <ul class=navbar-nav>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/"><i data-feather=home></i>Home</a>
+            <li class="nav-item "><a class=nav-link href="/"><i data-feather=home></i>Home</a>
             </li>
-            <li class="nav-item active"><a class=nav-link href="/previews/PR46/post/"><i data-feather=file-text></i>Blog</a>
+            <li class="nav-item active"><a class=nav-link href="/post/"><i data-feather=file-text></i>Blog</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/learn/"><i data-feather=book-open></i>Learn</a>
+            <li class="nav-item "><a class=nav-link href="/learn/"><i data-feather=book-open></i>Learn</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/cuda/">CUDA</a>
+            <li class="nav-item "><a class=nav-link href="/cuda/">CUDA</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/rocm/">ROCm</a>
+            <li class="nav-item "><a class=nav-link href="/rocm/">ROCm</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/oneapi/">oneAPI</a>
+            <li class="nav-item "><a class=nav-link href="/oneapi/">oneAPI</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/metal/">Metal</a>
+            <li class="nav-item "><a class=nav-link href="/metal/">Metal</a>
             </li>
-            <li class="nav-item "><a class=nav-link href="/previews/PR46/other/">Other</a>
+            <li class="nav-item "><a class=nav-link href="/other/">Other</a>
             </li>
         </ul>
     </nav>
@@ -175,7 +175,7 @@ <h2 id="new_features"><a href="#new_features" class="header-anchor">New features
     </footer>
 
     <!-- FEATHER -->
-    <script src="/previews/PR46/libs/feather/feather.min.js"></script>
+    <script src="/libs/feather/feather.min.js"></script>
     <script>feather.replace()</script>
 
     <!-- GOOGLE ANALYTICS -->

post/index.html

@@ -141,42 +141,50 @@ <h1>
   <a href=/post/index.xml><i data-feather=rss></i></a>
 </h1>
 
+<p>
+  <a class=font-125 href="/post/2024-09-18-cuda_5.5/">
+    CUDA.jl 5.5: Maintenance release
+  </a>
+  <br>
+  <i data-feather=calendar></i>
+  <time datetime=2024-9-18>Sep 18, 2024</time><br>
+</p>
 <p>
   <a class=font-125 href="/post/2024-05-28-cuda_5.4/">
     CUDA.jl 5.4: Memory management mayhem
-  </a><span>&nbsp;&#8599;</span>
+  </a>
   <br>
   <i data-feather=calendar></i>
   <time datetime=2024-5-28>May 28, 2024</time><br>
 </p>
 <p>
   <a class=font-125 href="/post/2024-05-24-oneapi_1.5/">
     oneAPI.jl 1.5: Ponte Vecchio support and oneMKL improvements
-  </a><span>&nbsp;&#8599;</span>
+  </a>
   <br>
   <i data-feather=calendar></i>
   <time datetime=2024-5-24>May 24, 2024</time><br>
 </p>
 <p>
   <a class=font-125 href="/post/2024-04-26-cuda_5.2_5.3/">
     CUDA.jl 5.2 and 5.3: Maintenance releases
-  </a><span>&nbsp;&#8599;</span>
+  </a>
   <br>
   <i data-feather=calendar></i>
   <time datetime=2024-4-26>Apr 26, 2024</time><br>
 </p>
 <p>
   <a class=font-125 href="/post/2023-11-07-cuda_5.1/">
     CUDA.jl 5.1: Unified memory and cooperative groups
-  </a><span>&nbsp;&#8599;</span>
+  </a>
   <br>
   <i data-feather=calendar></i>
   <time datetime=2023-11-7>Nov 7, 2023</time><br>
 </p>
 <p>
   <a class=font-125 href="/post/2023-09-19-cuda_5.0/">
     CUDA.jl 5.0: Integrated profiler and task synchronization changes
-  </a><span>&nbsp;&#8599;</span>
+  </a>
   <br>
   <i data-feather=calendar></i>
   <time datetime=2023-9-19>Sep 19, 2023</time><br>

post/index.xml

@@ -18,6 +18,39 @@
       type="application/rss+xml" />
 
 
+<item>
+  <title><![CDATA[CUDA.jl 5.5: Maintenance release]]></title>
+  <link>https://juliagpu.org/post/2024-09-18-cuda_5.5/index.html</link>
+  <guid>https://juliagpu.org/2024-09-18-cuda_5.5/</guid>
+  <description><![CDATA[CUDA.jl 5.5 is a minor release that comes with a couple of small improvements and new features.]]></description>  
+
+  <content:encoded><![CDATA[
+<p>CUDA.jl 5.5 is a minor release that comes with a couple of small improvements and new features.</p>
+<p>The only important change is that the minimal required Julia version has been bumped to 1.10, in anticipation of it becoming the next LTS release.</p>
+<h2 id="new_features">New features</h2>
+<ul>
+<li><p>Support for the upcoming Julia 1.11 release has been added, as well as for <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2461">CUDA 12.6 &#40;Update 1&#41;</a>.</p>
+</li>
+<li><p>Launch overhead has been reduced by <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2472">avoiding double argument conversions</a>. Note that this does not apply to kernels that are obtained using <code>@cuda launch&#61;false</code>.</p>
+</li>
+<li><p>CUSOLVER&#39;s dense wrappers have been improved by <a href="https://github.com/bjarthur">Ben Arthur</a>, <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2465">now caching workspace buffers</a>. This should greatly reduce the number of allocations needed for repeated calls.</p>
+</li>
+<li><p><a href="https://github.com/amontoison">Alexis Montoison</a> has improved the CUSPARSE wrappers, adding <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2489">conversions between sparse vectors and sparse matrices</a> that enable <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2488">a version of <code>gemv</code></a> which preserves sparsity of the inputs.</p>
+</li>
+<li><p>CUDA.jl&#39;s CUFFT wrappers <a href="https://github.com/JuliaGPU/CUDA.jl/pull/2430">now support <code>Float16</code></a>, thanks to <a href="https://github.com/eschnett">Erik Schnetter</a>.</p>
+</li>
+</ul>
+]]></content:encoded>
+
+  <pubDate>Wed, 18 Sep 2024 00:00:00 +0000</pubDate>  
+
+
+  <atom:author>
+    <atom:name>Tim Besard</atom:name>
+  </atom:author>
+
+</item>
+
 <item>
   <title><![CDATA[CUDA.jl 5.4: Memory management mayhem]]></title>
   <link>https://juliagpu.org/post/2024-05-28-cuda_5.4/index.html</link>