High results fluctuation in repeating experiments with same setup

I am using the command CUDA_VISIBLE_DEVICES=0,1 python3 SpecForge/benchmarks/bench_eagle3.py --model-path Qwen/Qwen3-8B --speculative-draft-model-path path/to/model --port 30000 --trust-remote-code --mem-fraction-static 0.8 --tp-size 1 --attention-backend fa3 --config-list 1,6,10,32 --benchmark-list mtbench --dtype bfloat16 to test 2 methods, however the first time I tested, baselineA results in TPS 106, baselineB results in TPS 100, the second time, baselineA results in TPS93, baseline B results in TPS 107, the third and later time I tested, both baselines are TPS 93......I don't think the randomness in the experiment is the cause, and the GPU and parameters for each experiment is all the same, does anyone faced the same issue or know about possible reasons?

Start with SpecForge/benchmarks/bench_eagle3.py and reproduce the repeated command using the stated model, GPU settings, and benchmark configuration. Compare the TPS results across repeated runs and trace the benchmark setup or runtime state that differs. Done means identifying a reproducible cause for the fluctuation or documenting the conditions under which it occurs.