Thank you ReduceMesh and McNeel

To whomever (maybe @Jussi_Aaltonen?): Thanks for ReduceMesh and it being multithreaded. Just reduced a 75mill poyfaces scan to a more managable 10mill without any problems. One question: The “preparing to reduce mesh” took an insane amount of time (40 minutes or so), while the actual reduction took only 15-20 minutes or so. Is this to be expected? I’m running a dual Xeon 2.6 with 96GB of RAM and a RTX4000 GPU - and neither CPU, GPU nor RAM saw more than sporadic activity during the prep phase.
TIA, Jakob