Showing the true power of the DGX Sparks : The AI Cookshow.
Two Sparks. One night. A model fleet wrote a playable roguelike from a single prompt, and then one of the models sat down and played its own game. No API. Zero cloud tokens. The setup: - Spark 1 runs Qwen 3.8 27B dense
I’m testing local llms for task analysis in Levelyss.
Same task: “Gain 500 followers on X” Gemma 4 12B - 45 XP - High importance - Skills: Marketing, Writing, Networking - Category: Work - 3.4s Gemma 4 26B - 35 XP - Medium importance - Skills: Marketing, Communication,
How much is the Doge in the window?
Gemini $0.09, GPT Luna $0.02, Muse Spark $0.11, Qwen (3.8:27b) $0.00. Nervos Network visual transaction counter. The pup barks when new blocks come in, a bark per transaction. ThreeJS here incorporating a locally generated GLB made with trellis
Also, working on a banger loop to run qwen 3.8 locally for hours to make 3D games on godot.
Took me all week but I think I'm finally getting this skill file set up correctly: the below ran for 7 hours on qwen 3.8b 27b on my 256gb studio (pretty buns for 7 hours imo) Would
You have reached the end of the archive
All of qwen38