<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Semih Güreşçi</title>
    <link>https://semihguresci.com/blog/</link>
    <description>Graphics and GPU engineer building real-time rendering systems in C++ and Vulkan, from mobile XR to desktop graphics.</description>
    <language>en</language>
    <atom:link href="https://semihguresci.com/feed.xml" rel="self" type="application/rss+xml" />
    
    <item>
      <title>GPU memory experiments: a guide to the 45-study series</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-series/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-series/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>A nine-part tour of 45 Vulkan compute experiments, from dispatch timing and data layout to subgroup algorithms and rendering workloads.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 40–45: queues, fusion, and reproducibility</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-40-45-queues-fusion-and-reproducibility/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-40-45-queues-fusion-and-reproducibility/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Persistent task claims, portable aggregates, dispatch fusion, layout sensitivity, and a reproducible single-GPU access signature.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 35–39: occupancy and rendering data</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-35-39-occupancy-and-rendering-data/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-35-39-occupancy-and-rendering-data/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>What shared-memory declarations, node records, frame remapping, visibility counters, and tiled light assignment actually measure.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 30–34: subgroups, transpose, and radix sort</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-30-34-subgroups-transpose-and-radix-sort/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-30-34-subgroups-transpose-and-radix-sort/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Portable subgroup indexing, block-local primitives, tiled transpose, and the cost of a complete GPU radix sort.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 26–29: alignment, heaps, and shared memory</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-26-29-alignment-heaps-and-shared-memory/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-26-29-alignment-heaps-and-shared-memory/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Four Vulkan studies separate contiguous offsets, access permutations, staging costs, and shared-memory stride from the hardware mechanisms their names suggest.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 21–25: reduction, scan, and compaction</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-21-25-reduction-scan-and-compaction/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-21-25-reduction-scan-and-compaction/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Reduction, prefix sums, histograms, compaction, and spatial binning show why ordering contracts and auxiliary traffic belong beside GPU timings.</description>
    </item>
    
    <item>
      <title>GPU memory experiments 16–20: tiling, live state, and barriers</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-16-20-tiling-registers-and-barriers/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-16-20-tiling-registers-and-barriers/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Five Vulkan experiments connect shared-memory reuse, tile size, temporary-state scheduling, branch patterns, and synchronization to measured work.</description>
    </item>
    
    <item>
      <title>GPU memory experiments, part 3: addresses matter as much as bytes</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-11-15-access-patterns-and-reuse/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-11-15-access-patterns-and-reuse/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Experiments 11–15 explore stride, gather, atomic scatter, scheduled reuse, and bandwidth curves while keeping their different address and validation contracts visible.</description>
    </item>
    
    <item>
      <title>GPU memory experiments, part 2: layout, padding, and precision</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-06-10-layout-padding-and-precision/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-06-10-layout-padding-and-precision/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Experiments 06–10 compare matched particle updates, record footprints, and manually packed formats—and expose why a shared correctness oracle can miss a conversion defect.</description>
    </item>
    
    <item>
      <title>GPU memory experiments, part 1: dispatch is part of the measurement</title>
      <link>https://semihguresci.com/blog/gpu-memory-experiments-01-05-dispatch-and-measurement/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-experiments-01-05-dispatch-and-measurement/</guid>
      <pubDate>Fri, 23 Oct 2026 00:00:00 +0000</pubDate>
      <description>Experiments 01–05 establish what a dispatch, a byte count, and a contiguous baseline actually mean before comparing GPU memory layouts.</description>
    </item>
    
    <item>
      <title>AoSoA and blocked layouts: measuring the tradeoffs</title>
      <link>https://semihguresci.com/blog/gpu-memory-access-benchmark-experiment-07-aosoa-blocked-layouts/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-access-benchmark-experiment-07-aosoa-blocked-layouts/</guid>
      <pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate>
      <description>A comparison of AoS, SoA, and blocked AoSoA layouts, including the performance and integration tradeoffs.</description>
    </item>
    
    <item>
      <title>GPU memory benchmarks: the first five experiments</title>
      <link>https://semihguresci.com/blog/gpu-memory-access-benchmark-experiments-01-05/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-access-benchmark-experiments-01-05/</guid>
      <pubDate>Wed, 20 May 2026 00:00:00 +0000</pubDate>
      <description>Five experiments establish a baseline for dispatch overhead, workgroup size, contiguous bandwidth, and indexing costs.</description>
    </item>
    
    <item>
      <title>AoS vs SoA: measuring GPU memory access</title>
      <link>https://semihguresci.com/blog/gpu-memory-access-benchmark-experiment-06-aos-vs-soa/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/gpu-memory-access-benchmark-experiment-06-aos-vs-soa/</guid>
      <pubDate>Wed, 20 May 2026 00:00:00 +0000</pubDate>
      <description>A measured comparison of AoS and SoA particle layouts, with an explanation of how field access affects GPU memory traffic.</description>
    </item>
    
    <item>
      <title>Building a Vulkan 1.4 scene renderer</title>
      <link>https://semihguresci.com/blog/announcing-vulkan-scene-renderer/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/announcing-vulkan-scene-renderer/</guid>
      <pubDate>Fri, 01 May 2026 00:00:00 +0000</pubDate>
      <description>Forward and deferred rendering, scene inspection, and GPU profiling in a C++23 and Vulkan 1.4 scene engine.</description>
    </item>
    
    <item>
      <title>Measuring Vulkan workloads with vk-bench</title>
      <link>https://semihguresci.com/blog/vk-bench-first-look/</link>
      <guid isPermaLink="true">https://semihguresci.com/blog/vk-bench-first-look/</guid>
      <pubDate>Mon, 06 Apr 2026 00:00:00 +0000</pubDate>
      <description>A small Vulkan benchmark for comparing CPU and GPU timings across controlled graphics and compute workloads.</description>
    </item>
    
  </channel>
</rss>
