Also, Datacenter GPUs only have passive cooling, presumably to allow for more Cuda cores. I get that what I have is older tech but for the cost of one H100 GPU I can have 60 of these servers (600 GPUs) plus some change to pay for CoLo fees.
For some of these, you could obviously pay with your time. Sometimes that time is very valuable, and sometimes you have a lot to spare.
Also, the amount of VRAM (3x)... Sometimes having too little of it means having to re-write the program, or it could dramatically impact the speed. Similarly for bandwidth (8x). But, again, if the kind of workload you have isn't constrained by either, then you could probably win by running on more smaller / older GPUs.