The problem with your benchmark is the calculation moves into floating point exception handling very early on.

If you avoid that by adding logs, using threads improve performance by 65%:

#! perl -slw
use strict;
use threads;
use Time::HiRes qw[ time ];

{
    my $start = time;
    my $t = 0;
    $t += log( $_ ) for 230000..900000;
    printf "%.15f\n", $t;;

    printf "Took %.6f\n", time() - $start;
}

{
    my $start = time;

    my $thr = async {
        my $t = 0;
        $t += log( $_ ) for 230000..565000;
        return $t;
    };
    my $t = 0;
    $t += log( $_ ) for 565001..900000;

    $t += $thr->join;

    printf "%.15f\n", $t;;

    printf "Took %.6f\n", time() - $start;
}


__END__
C:\test>junk32
8829606.110849546300000
Took 0.132204
8829606.110849652400000
Took 0.080650
[download]