Tag: local-ai

All the articles with the tag "local-ai".

• Benchmarking Qwen 27B and ComfyUI on One RTX 5090

  Published: 12 Mar, 2026

  • 5 min read

  I benchmarked several memory and cache combinations to see which RTX 5090 setup gave the best balance between image speed and VRAM safety.

• The RTX 5090 Setup That Let Me Run Qwen and ComfyUI Together

  Published: 12 Mar, 2026

  • 3 min read

  After testing several combinations on one RTX 5090, I landed on a setup that kept Qwen loaded while still generating images safely.