https://67ailab.com/tags/speculative-decoding/ Recent content in Speculative Decoding on 67AI Lab Hugo en-us Tue, 18 Aug 2026 23:11:08 +0000 https://67ailab.com/posts/qwen38-27b-dgx-spark-mtp-speedup/ Tue, 18 Aug 2026 23:11:08 +0000 https://67ailab.com/posts/qwen38-27b-dgx-spark-mtp-speedup/ <h2 id="deploying-qwen38-27b-on-a-dgx-spark-with-and-without-mtp">Deploying Qwen3.8-27B on a DGX Spark, with and without MTP</h2> <p>A dense 27B model on a DGX Spark generates about 11 tokens per second. That is not a bug, a bad build, or a thermal problem — it is arithmetic, and you can predict it before you download the weights.</p> <p>Then you turn on one setting and get 29 t/s, with identical output quality.</p> <p>Here is the whole story, measured on the box.</p>