AIXDAabout 2 hours ago
NAS runs AI at 5 tokens per second
1 min read
Ollama on a UGREEN DXP4800 Pro NAS runs Qwen3 4B at 5.4 tokens per second and Gemma 3 4B at 3.2, using CPU-only inference. The setup handles note formatting and Markdown cleanup well, though complex troubleshooting can take over 17 minutes.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
