https://cageyv.dev/series/llm/Recent content in LLM on Vladimir SamoylovHugoenFri, 06 Feb 2026 00:00:00 +0000https://cageyv.dev/posts/inf2-trn3-vllm-neuron-experiment/Fri, 06 Feb 2026 00:00:00 +0000https://cageyv.dev/posts/inf2-trn3-vllm-neuron-experiment/<p>I wanted a simple, repeatable way to answer a practical question:</p> <blockquote> <p>“If we had to run a private model on AWS, what’s the operational reality on Inf2 / Trn3?”</p></blockquote> <p>This post documents one small <strong>hands-on</strong> experiment (Inf2) plus the exact follow-up checks I’d run on <strong>Trn3</strong>.</p> <h2 id="experiment-setup-inf2"> Experiment setup (Inf2) <a class="heading-link" href="#experiment-setup-inf2"> <i class="fa-solid fa-link" aria-hidden="true" title="Link to heading"></i> <span class="sr-only">Link to heading</span> </a> </h2> <p>This was a quick “can we get it working end-to-end?” test, not a full benchmark suite.</p>