Hi choroba,

I tried using multiple cores for your demonstration. Perl has this capability. So little effort and practically transparent. :)

#!/usr/bin/perl
use warnings;
use strict;
use feature qw{ say };
use MCE::Util;
use MCE::Map max_workers => MCE::Util::get_ncpu();

sub collatz {
    my ($start) = @_;
    my @seq = $start;
    push @seq, ($seq[-1] / 2, 3 * $seq[-1] + 1)[$seq[-1] % 2]
        while $seq[-1] != 1;
    return @seq
}

my @sizes = mce_map_s { [$_, scalar collatz($_)] } 1, 1e6;

say "@$_" for reverse +(sort { $b->[1] <=> $a->[1] } @sizes)[0 .. 19];
[download]

Before and after on a 6 core machine with hyper-threading - 12 logical cores.

Serial    48.170 seconds
Parallel   9.361 seconds
[download]

Regards, Mario