llamaperf
Submit

NVIDIA P102-100 vs RTX 5090

For running local LLMs · 65 reports across 11 models

Side A
NVIDIA P102-100
Vendor
nvidia
VRAM
10GB
Memory
Discrete
Side B
RTX 5090
Vendor
nvidia
VRAM
32GB
Memory
Discrete

Highest reported generation speed by model family

These maxima come from different community setups. Model size, quantization, GPU count, offloading and concurrency may differ. They do not establish which GPU is faster for the same workload. Check the comparison conditions →

ModelNVIDIA P102-100RTX 5090
Qwen3.8up to 2400B—220.0n=34
DeepSeek V4 Flashup to 285B—120.0n=10
Qwen3.6up to 35B23.5n=1215.1n=6
Gemma 4up to 31B—578.0n=4
Museup to 30B—253.0n=4
GLM-5.2up to 744B——n=1
GLM-5.3—24.5n=1
Kimi K2.5——n=1
Qwen2.5up to 32B——n=1
Qwen3up to 235B—159.0n=1
Qwen3.5up to 122B——n=1

More comparisons