Qwen3.6 35B (3B active)
on 3× NVIDIA RTX Pro 6000 Blackwell · llama.cpp · 32,768 ctx
Oct 6, 2026
Use cases
summarization
Summary
User reports a prompt-processing regression on Qwen3.6-35B-A3B, dropping from 13816.18 tokens per second to 6858.24 tokens per second between llama.cpp builds 4ebdf2c74 and bed0a8566.
Setup is llama.cpp with Q4_K_M GGUF, 32768 context, -ngl 99 across 3x RTX Pro 6000 Blackwell (96 GB each) on a Threadripper PRO 9975WX with 256 GB RAM.
The regression appears in llama-completion and llama-server but not in llama-bench, where both builds measure about 15600 t/s at pp18606.