AI BriefingRTX 5070 Ti 44 t/s
·2026.04.24 20:02
Key point
On an RTX 5070 Ti 16GB and 32GB RAM, Qwen3.6-35B-A3B Q8_0 achieved 44 t/s.
Details
Running unsloth/Qwen3.6-35B-A3B-GGUF Q8_0 on an RTX 5070 Ti 16GB and 32GB DDR5 RAM setup achieved 44 t/s.
- Model size: 36.9 GB
- Context: 128K
- LM Studio settings:
- GPU Offload: 40
- Offload MoE Experts to CPU: 26
- Try mmap: on
- K cache: Q8_0
- V cache: Q8_0
The poster mentioned that llama.cpp would be better.
