The IO subsystem on Apple Silicon is different and won't scale nicely just by using all amount of available cores. Instead it seems best to only
use as many threads as performance cores are present on the system - otherwise the performance might actually get worse while using more power.
On all other systems, the default number of threads did not change.
Please note that for optimial performance one would need an arm build on MacOS, currently provided is only intel builds. To do that, on an Apple Silicon mac, run cargo install dua-cli.