AnalysisAI Models
10 hours ago
User runs Qwen3.6-27b at 55 tok/s on 4x RTX 5060 Ti for $1800
$1800 in GPUs (four RTX 5060 Ti 16GB) runs Qwen3.6-27b FP8 at 55 tok/s with 262K context using BF16 KV cache. The setup uses P2P communication between cards. Achieves inference for a single user.