Hello,

Today I tried to optimize slowest part of my program. That is inside of nested loops:

#!/usr/bin/perl

use warnings;
use strict;

$\ = $/;

srand( 1 );

my $n = 2e3;

@_ = map int rand 1e5, 1 .. $n;

print "@_[ 0 .. 3 ] ...";

my @max;

push @max, ( sort { $b <=> $a } @_ )[ 0 ];

my @sums = @_;

for my $i ( 1 .. $n - 1 ){
    
    my $max = -~0;
    for my $j ( $i .. $n - 1 ){
        
        # HOT-SPOT {
        $sums[ $j - $i ] += $_[ $j ];
        $sums[ $j - $i ] > $max and $max = $sums[ $j - $i ];
        # HOT-SPOT }
        
        }
    push @max, $max;
    }

print "Finished";
[download]

I changed the HOT-SPOT lines to:

        ( $sums[ $j - $i ] += $_[ $j ] ) > $max and $max = $sums[ $j -
+ $i ];
[download]

...and my code started to work ~30 percent faster. I think it is good increase in performance but a slight decrease in readability.

I executed these code dozens of times to see how fast do they work. I observed that both programs work in not constant time and about every 10th run program occasionally consume >1.5x time than usually. I observed it on perl v5.30 and v5.32, but not on v5.14. Any ideas why every ~10th run any of these programs runs much slower?