{"author_name":"sandeepknayak","title":"How We Cut GPU Inference Cost by ~75%: Async, Autoscaled Diffusion on GKE","blog_title":"Gaudiy Tech Blog","html":"<iframe src=\"https://hatenablog-parts.com/embed?url=https%3A%2F%2Ftechblog.gaudiy.com%2Fentry%2F2026%2F07%2F31%2F203039\" title=\"How We Cut GPU Inference Cost by ~75%: Async, Autoscaled Diffusion on GKE - Gaudiy Tech Blog\" class=\"embed-card embed-blogcard\" scrolling=\"no\" frameborder=\"0\" style=\"display: block; width: 100%; height: 190px; max-width: 500px; margin: 10px 0px;\"></iframe>","height":"190","version":"1.0","url":"https://techblog.gaudiy.com/entry/2026/07/31/203039","provider_url":"https://hatena.blog","provider_name":"Hatena Blog","description":"By Sandeep Kumar Nayak \u2014 AI Engineer at Gaudiy A technical deep-dive on how a single oversized GPU turned into ~90% of a bill, and the four changes that brought it back down. Framing note. This is a sanitized write-up of a real cost-optimization project. Service names, cluster names, and currency fi\u2026","type":"rich","width":"100%","categories":[],"blog_url":"https://techblog.gaudiy.com/","image_url":"https://cdn-ak.f.st-hatena.com/images/fotolife/g/gaudiy/20260731/20260731194641.png","published":"2026-07-31 20:30:39","author_url":"https://blog.hatena.ne.jp/sandeepknayak/"}