in reply to does perl have branch prediction

:P Whats happened when you tried and measured to see if its faster?

Awww:D

Replies are listed 'Best First'.
Re^2: does perl have branch prediction
by david2008 (Scribe) on Jan 29, 2014 at 06:38 UTC
    Good point Here is my code
    use strict; use warnings; use Benchmark; use List::Util qw /shuffle/; use feature 'say'; my $n = 10**7; my @sorted = 1..$n; my @random = shuffle @sorted; my $t0 = Benchmark->new; say sum_this(\@sorted); my $t1 = Benchmark->new; my $td1 = timediff($t1, $t0); say "sorted:",timestr($td1); my $t2 = Benchmark->new; say sum_this(\@random); my $t3 = Benchmark->new; my $td2 = timediff($t3, $t2); say "random:",timestr($td2); sub sum_this { my ($arr) = @_; my $sum; foreach (my $i=0;$i<=$#$arr;$i++) { my $x = $arr->[$i]; if($x>$n/2) { $sum+= $x; } } return $sum; }
    The output is
    37500002500000
    sorted: 6 wallclock secs ( 5.61 usr +  0.00 sys =  5.61 CPU)
    37500002500000
    random: 6 wallclock secs ( 5.81 usr +  0.00 sys =  5.81 CPU)
    
    There is a difference but not as dramatic as in java or c++ (from the stack overflow post) What is the correct interpretation of the results?