I’m already running make in -j8 and running multiple invocations of clang in parallel, why would I want each one of them using up 800% CPU? Effort spent making

I’m already running make in -j8 and running multiple invocations of clang in parallel, why would I want each one of them using up 800% CPU? Effort spent making a single invocation of clang saturate all my CPU cores is misguided IMO, and the wrong place to be doing parallelization.