Using perl script to fill up hard drive - can this be done more efficiently?

I have written a Perl script (Strawberry) to fill up some old hard drives with some semi-random numbers a couple of times before sending them to recycling.
The script creates a ~ 100mb file with random 0's and 1's which is then written 10 times into a file.
This process is then repeated depending on hard disk size (in the example only 3 times for testing purposes).

Currently, its taking around 36 seconds per file.
My question is: what changes can I make to the script to speed this up?

Here is the script in full:

my $master = "c:/programme/tools/filluphd/master.txt";
my $rd;

my $log = "c:/programme/tools/filluphd/log.log";

my $file = "f:/file_";

#create master
open (MASTER, ">", $master);

for ($i = 0; $i < 100000000; $i++)
{
    $rd = int(rand(2)); 
    print MASTER $rd;
}

close (MASTER);

open (LOG, ">", $log);
print LOG "start: ".localtime()."\n";
my $start_time = time();
my $random;
open (MASTER, "<", $master) or die "cannot open master: $!"."\n";
while (<MASTER>)
{
    $random .= $_;
}
close (MASTER);
for ($n = 0; $n <= 2 ; $n++) 
{
    my $file_start = time();
    my $filename = $file.$n.".txt";
    my $i = 0;
    open (FILE, ">", $filename);
    for ($i; $i <= 10; $i++)
    {
        print FILE $random;
    }
    close (FILE);
    my $file_end = time();
    my $file_runtime = $file_end - $file_start;
    print LOG "file ".$filename." took ".$file_runtime." seconds!"."\n";
}

my $end_time = time();
my $run_time = $end_time - $start_time;
print LOG "run time: ".$run_time." seconds"."\n";
print LOG "finish: ".localtime()."\n";
close (LOG);

I did some measuring and found out that the Case the HD is in is USB2-only. That means with about 30MB/s it is completely saturated and the windows tools show this. Speedwise, there is not much else to do.