boltbeam.fyi

Qwen3.8-Flash-Next speed on every GPU

5 measured runs across 3 GPUs, fastest first.

GPUQuantEnginetok/sRooflineShare
RTX 5090GSQ-RCO IQ1_Mllama.cpp + mods226.1528.543%
H200UD-Q4_K_XLllama.cpp100.4709.114%
RTX 5090GSQ-RCO IQ1_Mllama.cpp77.2537.114%
2× Radeon R9700GSQ-RCO IQ3_Sllama.cpp + MTP k=372.5––
2× Radeon R9700GSQ-RCO IQ3_Sllama.cpp39.4––

Download