Would be cool to have something like this in the standard library. This is getting close to being memory bound. In fact it will be memory bound on most amd64 ma

Would be cool to have something like this in the standard library. This is getting close to being memory bound. In fact it will be memory bound on most amd64 machines if you use more than 2 cores. It's kinda pointless to optimize much further, even if further optimization is possible.