<oembed><type>rich</type><version>1.0</version><title>utxo the webmaster 🧑‍💻 wrote</title><author_name>utxo the webmaster 🧑‍💻 (npub1ut…r50e8)</author_name><author_url>https://yabu.me/npub1utx00neqgqln72j22kej3ux7803c2k986henvvha4thuwfkper4s7r50e8</author_url><provider_name>njump</provider_name><provider_url>https://yabu.me</provider_url><html>I would definitely go for 512gb, probably will, gets you into that GLM class of models. But if I didn&#39;t already have sparks I would still grab a 256gb version today to run qwen and deepseek. Of course only do this if you think you can make money from it or have private workloads</html></oembed>