DiggTop ⌕

RTX 3090 vs RTX PRO 6000 Blackwell x2

Best sourced output tok/s per model (single-stream generation) — engine and version, quantization, context and a source link on every row. Hover a value for the quant.

RTX 309018 sourced runs · best 709.5 tok/s
RTX PRO 6000 Blackwell x23 verified runs · best 179.5 tok/s
ModelRTX 3090RTX PRO 6000 Blackwell x2
LFM2.5-1.2B-Instruct709.5 Q4_K_M—
Qwen3.5-0.8B-MTP476.1 Q6_K_XL—
NVIDIA-Nemotron-3.5-Lightning-30B-A3B353.7 NVFP4—
LFM2.5-2.6B305.6 Q4_K_M—
Qwen3.5-4B-MTP240.4 Q6_K_XL—
DeepSeek-V4-Flash-Vision-Exp—179.5 FP8
Qwen3.8-27B164.6 EXL3-3.5bpw—
Qwen3.5-9B-MTP164.3 Q6_K_XL—
GLM-5.3-Flash28.1 EXL3 3.05 bpw159.3 EXL3-3.0bpw
Qwen3.8-Flash-Next38.6 EXL3 2.50bpw—
Ornstein3.6-27B-MTP-NSC-ACE-SABER32.9 Q4_K_M—
DeepSeek-V4.1-Flash26.0 EXL3 3.0 bpw—

Head-to-head (models tested on both): RTX 3090 0 · RTX PRO 6000 Blackwell x2 1 · tied 0 · data as of Oct 6, 2026. Full rules: methodology.

Related reading: RTX 5090 vs 3090 · quantization · Value index · Full board

↑