I want to run benchmark with best utilization on my heterogeneous system

Hello Community, new here, need your assistance.

I have a heterogeneous system wherein I have two GPUs, GPU0-> RTX Quadro 8000 and GPU1-> RTX A6000.

I wanted to run some benchmarking code on this system. The problem is, I know that A6000 is more powerful (almost 2x) than Quadro 8000. This means that A6000 will not provide its peak performance when I do the benchmarking and thus I will get results either similar to using Quadro 8000 or just a little better.

So is there a way to achieve this, does anyone have a link to some resource which could guide me and your own guidance on making my other GPU i.e the A6000 work on double the data than Quadro 8000, while keeping the batch size the same. There;s also batch size constraint, I cant simply give a bigger batch to A6000 as both cards have same amount of GPU memory each. So what will be the best solution for this?

Thank You for your time and support!