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

my $seek = 'AGAGAGAA';

my $start = time();

my $file = shift;
open DNA, '<', $file or die $!;

my $size = -s( *DNA );

for my $o ( 1 .. $size - 100 ) {
    read( DNA, my $DNA, 100 );
    $DNA =~ tr[\n][]d;
    my $p =0;
    printf "%s found at position %d\n",
        $seek, $p while $p = 1+index $DNA, $seek, $p;

    seek( DNA, $o, 0 );
}

printf "Took %.3f seconds\n", time() - $start;

__END__
[ 1:32:07.24] C:\test>DNAsearch2 randDNA.txt | wc -l
1441865

[ 4:12:28.90] C:\test>